Ineffective Assistance in the Plea Process: Advice, Offers, and Consequences

By Elizabeth Franklin-Best, Esq., and Christopher Zoukis, JD, MBA

Can bad plea advice be challenged under § 2255?

Yes — bad plea advice is a §2255 claim. Under Hill v. Lockhart, 474 U.S. 52 (1985), you must show counsel’s advice fell below an objective standard of reasonableness and a reasonable probability that, but for the errors, you would not have pleaded guilty and would have insisted on going to trial. Unconveyed and rejected offers have their own tests.

Key takeaways on ineffective assistance in the plea process

• The prejudice test is different for pleas. Hill v. Lockhart holds that “the defendant must show that there is a reasonable probability that, but for counsel’s errors, he would not have pleaded guilty and would have insisted on going to trial.” That is the whole prejudice inquiry for a plea you accepted — not whether you would have won.
• You do not have to show you would have been acquitted. Jae Lee v. United States, 582 U.S. 357 (2017), rejected a per se rule that a defendant with no viable trial defense cannot show prejudice, because “the Hill v. Lockhart inquiry focuses on a defendant’s decisionmaking.”
• An offer your lawyer never told you about is a claim. Missouri v. Frye, 566 U.S. 134 (2012): “As a general rule, defense counsel has the duty to communicate formal prosecution offers to accept a plea on terms and conditions that may be favorable to the accused.”
• Contemporaneous evidence, not hindsight, wins these cases. Lee directs courts to “look to contemporaneous evidence to substantiate a defendant’s expressed preferences” rather than “post hoc assertions” — which is why the file, the emails, and the plea transcript matter more than an affidavit written years later.
• A plea colloquy is a barrier, not a wall. Blackledge v. Allison, 431 U.S. 63 (1977), holds that “[s]olemn declarations in open court carry a strong presumption of verity,” but also that “the barrier of the plea or sentencing proceeding record, although imposing, is not invariably insurmountable.”
• An appeal waiver obtained through bad advice is not enforceable. In Hunter v. United States, No. 24-1063 (U.S. June 18, 2026), the Court stated that “[a]n appeal waiver, again like the rest of the agreement, is not knowing and voluntary if it was the product of ineffective assistance of counsel.”

Can I file a §2255 motion if my lawyer’s plea advice was wrong?

Yes. A claim that your lawyer’s advice caused you to plead guilty is a Sixth Amendment ineffective-assistance claim, and it is raised in a motion under 28 U.S.C. §2255 filed in the district court that sentenced you. It is not raised by asking to withdraw the plea, because under Fed. R. Crim. P. 11(e), “[a]fter the court imposes sentence, the defendant may not withdraw a plea of guilty or nolo contendere, and the plea may be set aside only on direct appeal or collateral attack.” Once you are sentenced, the collateral attack is the vehicle — the three-stage withdrawal sequence, and why the Rule 11(e) cutoff routes a plea claim here, is set out at can you withdraw a guilty plea after sentencing. What a §2255 motion is and what it can reach is covered at what is a 2255 motion.

There is a doctrinal reason a plea claim has to be framed this way rather than as a complaint about something that happened earlier in the case. Tollett v. Henderson, 411 U.S. 258 (1973), holds that “[w]hen a criminal defendant has solemnly admitted in open court that he is in fact guilty of the offense with which he is charged, he may not thereafter raise independent claims relating to the deprivation of constitutional rights that occurred prior to the entry of the guilty plea.” A guilty plea, the Court said, “represents a break in the chain of events which has preceded it in the criminal process.” A bad search, a coerced statement, a defective indictment — as freestanding claims, those generally do not survive the plea.

But Tollett also states the exception that this entire page occupies: a defendant “may only attack the voluntary and intelligent character of the guilty plea by showing that the advice he received from counsel was not within the standards set forth in McMann.” And it goes further — “[c]ounsel’s failure to evaluate properly facts giving rise to a constitutional claim, or his failure properly to inform himself of facts that would have shown the existence of a constitutional claim, might in particular fact situations meet this standard of proof.” The suppressible search does not survive the plea on its own. Counsel’s failure to recognize it, and the effect that failure had on the decision to plead, can.

The governing standard for the lawyer’s conduct is the ordinary two-prong test from Strickland v. Washington, 466 U.S. 668 (1984) — deficient performance plus prejudice — which is set out in full at Strickland v. Washington. What changes in the plea setting is the second prong, and it changes enough that the two claims are not interchangeable.

What does Hill v. Lockhart require me to prove?

Hill v. Lockhart, 474 U.S. 52 (1985), is the case that carried Strickland into the plea context, and it is the starting point for every claim on this page. The Court held: “the two-part Strickland v. Washington test applies to challenges to guilty pleas based on ineffective assistance of counsel.” Then it rewrote the second half of that test for pleas:

“The second, or ‘prejudice,’ requirement, on the other hand, focuses on whether counsel’s constitutionally ineffective performance affected the outcome of the plea process. In other words, in order to satisfy the ‘prejudice’ requirement, the defendant must show that there is a reasonable probability that, but for counsel’s errors, he would not have pleaded guilty and would have insisted on going to trial.”

Read that carefully, because two things follow from it that trip up nearly every pro se motion we see.

First, the showing is about your decision, not about the verdict. The question is not whether a jury would have acquitted you. It is whether there is a reasonable probability you would have refused the deal and demanded a trial. Those are different questions with different evidence behind them.

Second, you have to actually say it. Hill lost on exactly this point. His lawyer misadvised him about his parole-eligibility date, and the Court found it “unnecessary to determine whether there may be circumstances under which erroneous advice by counsel as to parole eligibility may be deemed constitutionally ineffective assistance,” because “[p]etitioner did not allege in his habeas petition that, had counsel correctly informed him about his parole eligibility date, he would have pleaded not guilty and insisted on going to trial.” He also “alleged no special circumstances that might support the conclusion that he placed particular emphasis on his parole eligibility in deciding whether or not to plead guilty.” A motion that describes terrible advice and stops there is not a Hill motion. It is a Hill dismissal.

Hill also gave courts a bridge for one category of claim. Where the alleged failure is a failure to investigate or to discover exculpatory evidence, whether that failure “prejudiced” the defendant “will depend on the likelihood that discovery of the evidence would have led counsel to change his recommendation as to the plea,” which “will depend in large part on a prediction whether the evidence likely would have changed the outcome of a trial.” For an uninvestigated-defense claim, then, the strength of the defense does re-enter the analysis — through counsel’s recommendation, not through a direct prediction of acquittal.

The plea-stage prejudice table

The six fact patterns below are not variations on a theme. Each has its own controlling decision and its own required showing, and a motion that pleads the wrong one loses on the papers even when the underlying complaint is real. This is the table to bring to a first consultation.

Table 1. What you must prove, by plea-stage fact pattern

Your fact pattern	Controlling decision	The specific prejudice showing required	What the court will look for in the record
Counsel gave you wrong advice and you accepted a plea	Hill v. Lockhart, 474 U.S. 52 (1985)	A reasonable probability that, but for counsel’s errors, you “would not have pleaded guilty and would have insisted on going to trial”	An express allegation that you would have gone to trial; “special circumstances” showing you placed particular emphasis on the misadvised point
Counsel never conveyed a formal offer and it lapsed	Missouri v. Frye, 566 U.S. 134 (2012)	A reasonable probability that you would have accepted the offer, and that it would have been entered “without the prosecution’s canceling it or the trial court’s refusing to accept it”	The written offer and its expiration date; the file showing no transmittal; the later, worse disposition you actually received
Counsel’s advice caused you to reject an offer and you were convicted at trial	Lafler v. Cooper, 566 U.S. 156 (2012)	A reasonable probability that the offer would have been presented to the court, that the court would have accepted its terms, and that the conviction or sentence under the offer “would have been less severe” than what was imposed	Proof the offer existed and was rejected on counsel’s advice; the arithmetic gap between the offer and the judgment
Counsel misadvised you (or said nothing) about deportation	Padilla v. Kentucky, 559 U.S. 356 (2010); Jae Lee v. United States, 582 U.S. 357 (2017)	A reasonable probability you would have rejected the plea; no showing of a likely acquittal is required, but rejecting the plea must have been “rational under the circumstances”	Contemporaneous evidence that immigration status was determinative — plea-colloquy exchanges, counsel’s notes, emails, family ties, length of residence
Counsel misstated the Guidelines range or a mandatory minimum	Strickland/Hill framework; e.g. United States v. Stephens, No. 23-1464 (3d Cir. Oct. 15, 2025)	Advice “so incorrect and so insufficient that it undermine[d]” your ability to make an intelligent decision, plus the Hill showing	The plea agreement, the Rule 11 transcript, counsel’s written estimate, and the actual calculated range in the PSR
Counsel’s error cost you the chance to negotiate a different plea	Rodriguez-Penton v. United States, 905 F.3d 481 (6th Cir. 2018), and the circuits it follows	In those circuits, a reasonable probability that you “would have bargained for a more favorable plea” — no showing that you would have gone to trial	Comparable plea agreements in similar cases; evidence the government would have accepted an alternative charge

Table 1 is not a substitute for legal advice. Which row applies depends on facts that only a review of your own record can establish, and the last row is not the law in every circuit.

Applied Insight — Elizabeth Franklin-Best, Esq., Principal Attorney: The first thing we do with a plea-advice case is decide which row of that table it actually sits in, because the answer changes what we go looking for. An unconveyed-offer case is a documents case — we want the letter, the docket, the file. A misadvice case is a conversation case — we want everything written down anywhere near the conversation. A rejected-offer case is an arithmetic case — we want the offer’s terms and the judgment side by side. Motions that plead all four theories at once tend to prove none of them, because the record built for one is not the record the others need.

What if my lawyer never told me about a plea offer?

An offer your lawyer never passed along is one of the strongest fact patterns in this area, and it has its own Supreme Court case. Missouri v. Frye, 566 U.S. 134 (2012), holds: “As a general rule, defense counsel has the duty to communicate formal prosecution offers to accept a plea on terms and conditions that may be favorable to the accused.”

The facts are worth knowing because they are ordinary. Galin Frye was charged with driving with a revoked license. Because he had three prior convictions for the same offense, Missouri charged it as a class D felony carrying a four-year maximum. On November 15, 2007, the prosecutor sent Frye’s counsel a letter with two offers: plead to the felony with a three-year recommendation and ten days of “shock” time, or plead to a reduced misdemeanor with a 90-day recommendation. Both expired December 28. Counsel never told Frye either offer existed. Frye later pleaded guilty with no agreement at all and was sentenced to three years in prison — on a charge that could have been a misdemeanor with a 90-day recommendation.

Frye rests on a structural point: “The Sixth Amendment right to effective assistance of counsel extends to the consideration of plea offers that lapse or are rejected,” because plea bargaining is a critical stage of the proceeding. The Government’s argument that a later knowing and voluntary plea cures the earlier failure was rejected.

The prejudice showing, though, is more demanding than “he would have taken it.” Frye requires “a reasonable probability both that they would have accepted the more favorable plea offer had they been afforded effective assistance of counsel and that the plea would have been entered without the prosecution’s canceling it or the trial court’s refusing to accept it, if they had the authority to exercise that discretion under state law.” The Court explained why: “a defendant has no right to be offered a plea … nor a federal right that the judge accept it.” That second half is where these claims are won and lost, and it is why the Court in Frye did not simply grant relief — it vacated and remanded so the state court could address whether the prosecution might have withdrawn the offer in light of Frye’s new arrest a week before his hearing.

One naming point, because search engines conflate them: Missouri v. Frye has nothing to do with the “Frye standard” for admitting scientific evidence, which comes from an unrelated and much older decision. If you arrived here looking for the expert-testimony rule, this is not that case.

What if my lawyer told me to reject an offer?

Lafler v. Cooper, 566 U.S. 156 (2012), decided the same day as Frye, covers the mirror-image situation: the offer reached you, and counsel’s advice caused you to turn it down.

Anthony Cooper was charged in Michigan with assault with intent to murder and related counts after shooting Kali Mundy in the buttock, hip, and abdomen. The prosecution twice offered to dismiss some charges and recommend 51 to 85 months. Cooper admitted guilt in a communication with the court and said he was willing to accept — then rejected the offer on both occasions after his attorney convinced him the prosecution could not establish intent to murder because the victim had been shot below the waist. He went to trial, was convicted on all counts, and received a mandatory minimum of 185 to 360 months — a minimum, as the Court put it, “3½ times greater than he would have received under the plea.” Everyone conceded the advice was deficient; the only question was prejudice.

The Court’s answer is the test in row three of Table 1: a defendant “must show that but for the ineffective advice of counsel there is a reasonable probability that the plea offer would have been presented to the court (i.e., that the defendant would have accepted the plea and the prosecution would not have withdrawn it in light of intervening circumstances), that the court would have accepted its terms, and that the conviction or sentence, or both, under the offer’s terms would have been less severe than under the judgment and sentence that in fact were imposed.”

Lafler also answered the objection that a fair trial cures everything. It does not: “The Sixth Amendment requires effective assistance of counsel at critical stages of a criminal proceeding. Its protections are not designed simply to protect the trial.” Justice Scalia’s dissent argued the opposite — that the right “is not infringed unless counsel’s mistakes call into question the basic justice of a defendant’s conviction or sentence” — and lost. If you read a page that describes that position as the law, it is describing the dissent.

What is a “Lafler hearing,” and what is the remedy?

There is no rule creating a proceeding called a “Lafler hearing”; the phrase is shorthand for the evidentiary hearing Lafler contemplates on remand. The Court said that where the only advantage of the lost offer was a shorter sentence on the same charges, “the court may conduct an evidentiary hearing to determine whether the defendant has shown a reasonable probability that but for counsel’s errors he would have accepted the plea. If the showing is made, the court may exercise discretion in determining whether the defendant should receive the term of imprisonment the government offered in the plea, the sentence he received at trial, or something in between.”

Where resentencing is not enough — “if an offer was for a guilty plea to a count or counts less serious than the ones for which a defendant was convicted after trial, or if a mandatory sentence confines a judge’s sentencing discretion after trial” — the Court said “the proper exercise of discretion to remedy the constitutional injury may be to require the prosecution to reoffer the plea proposal,” after which “the judge can then exercise discretion in deciding whether to vacate the conviction from trial and accept the plea or leave the conviction undisturbed.” The Court added that in exercising that discretion “a court may take account of a defendant’s earlier expressed willingness, or unwillingness, to accept responsibility.”

Whether you get any hearing at all is its own fight, governed by §2255(b) and Rule 8 of the Rules Governing Section 2255 Proceedings; that question belongs to 2255 evidentiary hearing.

What if I was never told the plea would get me deported?

Padilla v. Kentucky, 559 U.S. 356 (2010), holds that “counsel must inform her client whether his plea carries a risk of deportation.” Jose Padilla, a native of Honduras, had been a lawful permanent resident for more than 40 years and served in the U.S. Armed Forces during the Vietnam War. He pleaded guilty to transporting a large amount of marijuana in his tractor-trailer in Kentucky. His counsel not only failed to warn him of the consequence but told him he “did not have to worry about immigration status since he had been in the country so long.” The Court held that “constitutionally competent counsel would have advised him that his conviction for drug distribution made him subject to automatic deportation.”

Padilla draws a line that matters enormously in practice. When the removal consequence is clear, counsel must give correct advice; when it is not, a general warning is enough. In the Court’s words: “When the law is not succinct and straightforward … a criminal defense attorney need do no more than advise a noncitizen client that pending criminal charges may carry a risk of adverse immigration consequences. But when the deportation consequence is truly clear, as it was in this case, the duty to give correct advice is equally clear.” The Court also held there is “no relevant difference ‘between an act of commission and an act of omission’ in this context” — saying nothing can be as deficient as saying something wrong.

Two limits belong in the same breath, and competitor pages routinely omit both.

Limit one — the Fed. R. Crim. P. 11 warning is not the same as advice. Rule 11(b)(1)(O) requires the court to advise a defendant “that, if convicted, a defendant who is not a United States citizen may be removed from the United States, denied citizenship, and denied admission to the United States in the future.” That is a generic may. It is not the case-specific analysis Padilla requires of counsel when the statute is clear, and Jae Lee is the proof: Lee received a version of that warning from the judge and still prevailed.

Limit two — Padilla is not retroactive. Chaidez v. United States, 568 U.S. 342 (2013), holds that “under the principles set out in Teague v. Lane, 489 U.S. 288 (1989), Padilla does not have retroactive effect.” Padilla was a “new rule” because it “first rejected the categorical approach” that had excluded advice about a conviction’s non-criminal consequences from Sixth Amendment scrutiny. The practical line is stark: if your federal conviction became final before March 31, 2010, Padilla does not help you. A conviction that became final afterward is governed by it. How a new decision does or does not reach an already-final case is the subject of retroactive Supreme Court decisions. The Padilla duty itself, the Chaidez bar, the finality-date computation it turns on, and the circuit split over pre-2010 affirmative-misadvice claims are covered in full at Padilla v. Kentucky.

Do I have to show I would have won at trial?

No — and this is the single most important thing on this page, because it is the argument the Government makes in nearly every one of these cases.

Jae Lee v. United States, 582 U.S. 357 (2017), squarely rejected “a per se rule that a defendant with no viable defense cannot show prejudice from the denial of his right to trial.” The Court gave two reasons. First, “categorical rules are ill suited to an inquiry that demands a ‘case-by-case examination’ of the ‘totality of the evidence.'” Second, and more fundamentally, the Government’s rule “overlooks that the Hill v. Lockhart inquiry focuses on a defendant’s decisionmaking, which may not turn solely on the likelihood of conviction after trial.”

The facts show why. Jae Lee came to the United States from South Korea in 1982, at thirteen. He built two restaurants in suburban Memphis and, in 35 years, never returned to Korea, living as a lawful permanent resident. In 2008 agents searched his house and found 88 ecstasy pills, three Valium tablets, $32,432 in cash, and a loaded rifle. He admitted the drugs were his. He was indicted on one count of possessing ecstasy with intent to distribute under 21 U.S.C. §841(a)(1). He repeatedly asked his attorney whether he would be deported; the attorney assured him he would not. He pleaded guilty and received a year and a day — and then learned he had pleaded to an “aggravated felony” under 8 U.S.C. §1101(a)(43)(B), which made his removal mandatory under 8 U.S.C. §1227(a)(2)(A)(iii). He had no real defense to the charge. He won anyway.

The Court’s reasoning is the part to carry away: “The decision whether to plead guilty also involves assessing the respective consequences of a conviction after trial and by plea. … When those consequences are, from the defendant’s perspective, similarly dire, even the smallest chance of success at trial may look attractive.” For Lee, “deportation after some time in prison was not meaningfully different from deportation after somewhat less time,” so he “would have rejected any plea leading to deportation in favor of throwing a ‘Hail Mary’ at trial.” The Court also disposed of Strickland‘s “lawless decisionmaker” language, explaining that the presumption of reliability applied to judicial proceedings “has no place where, as here, a defendant was deprived of a proceeding altogether.”

There is a limit, and it is real. Lee also holds: “Courts should not upset a plea solely because of post hoc assertions from a defendant about how he would have pleaded but for his attorney’s deficiencies. Rather, they should look to contemporaneous evidence to substantiate a defendant’s expressed preferences.” And Padilla had already added that “a petitioner must convince the court that a decision to reject the plea bargain would have been rational under the circumstances.” So the door is open, but you walk through it with documents, not with an affidavit.

Where the circuits disagree — and where they no longer do

Before Lee, the courts of appeals had split over whether the plea-prejudice inquiry was purely objective. Heard v. Addison, 728 F.3d 1170 (10th Cir. 2013), cataloged “a point that appears to have caused some confusion among the circuits; namely, whether the ‘prejudice’ inquiry contains a ‘subjective’ component,” contrasting the Fourth Circuit’s view that a petitioner’s statement “carries some probative value” with the Sixth Circuit’s statement that “[t]he [prejudice] test is objective, not subjective.” The Tenth Circuit’s own answer was a two-step: ask first whether going to trial “would have been objectively ‘rational under the circumstances'”; then, past that threshold, consider the individual defendant’s statements and conduct. Lee later confirmed the individualized focus while imposing the contemporaneous-evidence requirement, which is why post-2017 opinions read more consistently than the pre-2017 ones.

A second disagreement is live. Several circuits hold that prejudice can rest on the lost chance to negotiate a different plea, not only on a willingness to go to trial. In Rodriguez-Penton v. United States, 905 F.3d 481 (6th Cir. 2018), the Sixth Circuit held that a movant “may demonstrate prejudice if he can show that, had he known about the risk of adverse immigration consequences, he would have bargained for a more favorable plea,” and noted that the five circuits to have addressed the question — the Second, Fourth, Seventh, Ninth, and D.C. Circuits — had all reached the same conclusion. The opinion drew a dissent arguing that “trials are a matter of right, but plea negotiations are not.” Do not assume this theory is available in your circuit. It is a genuine variation in the law, and it is one of the reasons the choice of theory is not a matter of style.

Does the plea colloquy prevent me from claiming this?

It does not bar the claim, but it is the largest single obstacle in most of these cases, and any honest page has to say so.

Under Fed. R. Crim. P. 11(b), before accepting your plea the judge had to address you personally in open court and confirm you understood the charges, the maximum penalty, any mandatory minimum, the trial rights you were giving up, and the court’s obligation to calculate and consider the Guidelines range. Under Rule 11(b)(2) the judge had to “determine that the plea is voluntary and did not result from force, threats, or promises (other than promises in a plea agreement).” Under Rule 11(b)(3) the judge had to find a factual basis. You answered those questions under oath, and the transcript exists.

Blackledge v. Allison, 431 U.S. 63 (1977), describes exactly how much weight that carries: “the representations of the defendant, his lawyer, and the prosecutor at such a hearing, as well as any findings made by the judge accepting the plea, constitute a formidable barrier in any subsequent collateral proceedings. Solemn declarations in open court carry a strong presumption of verity.” And it states the consequence: “The subsequent presentation of conclusory allegations unsupported by specifics is subject to summary dismissal, as are contentions that in the face of the record are wholly incredible.”

But Blackledge also holds the other half, and the other half is why these cases are filed: “the barrier of the plea or sentencing proceeding record, although imposing, is not invariably insurmountable.” The Court refused to “adopt a per se rule excluding all possibility that a defendant’s representations at the time his guilty plea was accepted were so much the product of such factors as misunderstanding, duress, or misrepresentation by others as to make the guilty plea a constitutionally inadequate basis for imprisonment.”

Two practical points follow. First, most colloquies ask whether anyone made promises outside the agreement — not what your lawyer privately predicted, and not whether the advice you got was accurate. A claim that counsel told you the Guidelines would land at 60 months does not necessarily contradict a “no promises” answer. Second, Jae Lee shows that the colloquy can cut for you. Lee’s own colloquy became his best evidence: when the judge warned that a conviction “could result in your being deported” and asked whether that affected his decision, Lee answered “Yes, Your Honor,” then said “I don’t understand” and turned to his attorney — who told him it was a “standard warning.” That exchange is contemporaneous evidence that immigration status was determinative, and it is in the transcript.

Applied Insight — Christopher Zoukis, JD, MBA, Managing Director: People assume a plea-advice case is built on what they remember. It is built on what can be produced. The materials that move these motions are almost always things that already exist and are simply scattered — the plea transcript and the sentencing transcript, the plea agreement, the discovery cover letters, the lawyer’s written Guidelines estimate, the visiting-room and phone logs showing when meetings happened, the emails or letters between a family member and counsel, and the client file itself, which you are entitled to request. When we open a case, we ask for those before anyone drafts a word, because the timeline they build is what separates a claim a court will hear from one it dismisses as conclusory. Requesting counsel’s file also has legal consequences for privilege, and the legal side of that decision is handled by the firm’s attorneys.

What about wrong advice on the Guidelines or a mandatory minimum?

This is the most common plea-advice complaint in federal court and the hardest of the fact patterns to win, because a great deal of it is prediction — and a prediction that turns out wrong is not automatically deficient performance.

The line the cases actually draw is between an inaccurate forecast and an incorrect statement of law. In United States v. Stephens, No. 23-1464 (3d Cir. Oct. 15, 2025), the Third Circuit put it this way: “familiarity with the structure and basic content of the Guidelines (including the definition and implications of career offender status) has become a necessity for counsel who seek to give effective representation,” though counsel need not provide “a detailed exegesis of the myriad arguably relevant nuances of the Guidelines.” Performance is substandard where the advice is “so incorrect and so insufficient that it undermine[s the defendant’s] ability to make an intelligent decision about whether to accept the offer.” Applying that, the court reasoned that “any criminal defense attorney would have been aware that if the career offender provision applies, the resulting guideline range is still ‘advisory'” — so if counsel told the defendant the provision carried a mandatory minimum, “his performance was ‘unreasonable under prevailing professional norms.'” The court vacated the summary denial and remanded for a hearing, noting that the §2255 hearing standard “is not demanding” and that a movant “need not ‘prove’ anything” to get one.

Three practical filters come out of that, and they apply well beyond the Third Circuit:

1. A wrong number is weaker than a wrong rule. “I thought you’d land at 70 months” is a prediction. “The career-offender guideline is a mandatory minimum,” “this count doesn’t stack,” or “that prior doesn’t count” are statements of law that can be checked and can be wrong.
2. The gap has to matter. Because the Hill showing is about your decision, the difference between what you were told and what you got has to be large enough to plausibly have changed it. A three-month variance rarely is; the difference between a ten-year exposure and a mandatory life exposure obviously is.
3. Any additional prison time counts as prejudice at sentencing. That principle comes from the Strickland line rather than from the plea cases, and it is developed at Strickland v. Washington.

The written record usually decides these. Where the plea agreement and the Rule 11 colloquy both stated the correct statutory maximum and mandatory minimum, courts routinely hold the movant cannot have been misled about statutory exposure. Where the misadvice concerned a Guidelines classification never mentioned in either document, the record does not foreclose the claim — which is exactly the posture that produced the remand in Stephens.

Does my appeal waiver bar this claim?

Usually not, when the claim is that ineffective assistance made the plea itself invalid — and there is now Supreme Court language directly on the point.

In Hunter v. United States, No. 24-1063 (U.S. June 18, 2026), the Court’s principal holding concerned sentencing appeals: an agreement not to appeal a sentence “is unenforceable when it would result in a miscarriage of justice—meaning, when it would leave in place the kind of egregious error that would bring the judicial system into disrepute.” In the course of getting there, the Court stated the rule that matters here, citing Hill v. Lockhart: “An appeal waiver, again like the rest of the agreement, is not knowing and voluntary if it was the product of ineffective assistance of counsel. … So an appeal waiver tainted by ineffective assistance is not enforceable.” The Court noted it was not addressing “the circumstances in which an appeal waiver might be invalidly obtained, as through fraud or coercion.”

That aligns with what the courts of appeals had already been doing. The Fifth Circuit rule the Court reviewed in Hunter recognized two exceptions to enforceability — where “the waiver itself was ‘tainted by ineffective assistance of counsel'” and where the sentence “exceed[ed] the statutory maximum” — and the Hunter opinion observes that “[e]very Circuit, including the Fifth,” has recognized the ineffective-assistance principle.

Two cautions belong with that. First, the waiver’s reach depends on the claim, not on the label. A claim that counsel’s advice made the plea and the waiver unknowing attacks the waiver’s validity. A claim that counsel performed badly at sentencing on a point unrelated to the plea’s validity often falls squarely inside a waiver’s terms. Second, the Supreme Court has not resolved how Hunter applies to collateral-attack waivers specifically. In Jones v. United States, No. 25-524 (U.S. June 30, 2026), the Court denied certiorari; Justice Sotomayor filed a statement encouraging lower courts to consider Hunter‘s reasoning when deciding whether to enforce a waiver against a defendant challenging the validity of his conviction, and noting that Hunter itself arose “in the sentencing context.” A statement respecting a denial of certiorari is not a holding and creates no rule. What waivers say, what they cover, and what survives them is the subject of appeal waiver.

What is the remedy if I win?

The remedy depends on which fact pattern you proved, and it is almost never simply “you go free.”

If you proved you would not have pleaded guilty — the Hill pattern — the conviction and sentence are vacated and the case returns to its pre-plea posture. That means the original charges are live again, including any counts the government dismissed in exchange for the plea. This is the point at which some clients decide not to file. A movant who pleaded to one count facing a five-count indictment needs to understand that winning restores the indictment.

If you proved a lost or rejected offer — the Frye and Lafler patterns — Lafler supplies a menu rather than a rule. Where the plea and the trial conviction involved the same charges, the court may resentence you to “the term of imprisonment the government offered in the plea, the sentence he received at trial, or something in between.” Where the lost offer involved less serious counts, or where a mandatory sentence limited the judge’s discretion after trial, the court may “require the prosecution to reoffer the plea proposal,” and then decide “whether to vacate the conviction from trial and accept the plea or leave the conviction undisturbed.” The judge retains discretion at every step, and may weigh “a defendant’s earlier expressed willingness, or unwillingness, to accept responsibility.”

If you are no longer in custody, §2255 is unavailable regardless of how strong the claim is — the statute requires custody. A person who has completed a sentence and is facing collateral consequences, including immigration consequences, has to look at a different writ; that route is covered at coram nobis.

What happens procedurally between filing and any of these outcomes — the government’s response, the record expansion, the ruling, and the timeline — is set out at what happens after filing a 2255.

How do you prove advice that was never written down?

Almost every claim on this page rests on a conversation with no transcript. Lee tells courts to look for “contemporaneous evidence,” and a movant who has none is usually left with a “post hoc assertion” — the exact thing Lee says courts should not upset a plea over. So the work is finding the contemporaneous record that already exists.

In practice, that record has five layers.

The court file. The plea agreement, the Rule 11 transcript, the presentence report and the objections to it, the sentencing transcript, and the judgment. These are the documents the court will read first, and they set the outer boundary of what you can plausibly claim you did not know.

The prosecution’s paper. Written plea offers, cover emails, extensions, and expiration dates. In an unconveyed-offer case, this is the case. Frye itself turned on a letter dated November 15 with a December 28 expiration.

Counsel’s file. You are generally entitled to your own file. It may contain the Guidelines worksheet, the offer, notes of jail visits, and correspondence. Requesting it has legal consequences for the attorney-client privilege, because filing an ineffectiveness claim generally waives privilege as to the subject matter of the claim — a trade-off that has to be made deliberately, with counsel, not by accident.

Third-party contemporaneous accounts. Letters and emails to family members written before sentencing, jail phone-call and visiting logs, and — for a Padilla or Lee claim — evidence of the ties that make removal the decisive consequence: length of residence, U.S.-citizen children, a business, military service.

The colloquy itself. As Lee demonstrates, a defendant’s own confusion on the record can be the strongest contemporaneous evidence in the file.

A claim assembled from those layers looks entirely different from one that says “my lawyer told me I would get five years.” The first is a set of exhibits with dates on them. The second is a memory, and Blackledge explains what courts do with memories that contradict a transcript.

When is the deadline, and what if the direct appeal is still running?

A §2255 motion is generally due within one year, and the clock most often starts when the judgment of conviction becomes final. Computing that date from the events in your own case — including what happens when you did not appeal, when you appealed and lost, and when you sought certiorari — is the job of 2255 statute of limitations. The complete inventory of every other clock running after a federal conviction is at post-conviction deadlines. If the one-year period has already run, the narrow doctrine that can sometimes save a late motion is at equitable tolling.

Two timing points are specific to plea-advice claims.

You do not have to raise ineffectiveness on direct appeal to preserve it. Massaro v. United States, 538 U.S. 500 (2003), holds that “an ineffective-assistance-of-counsel claim may be brought in a collateral proceeding under §2255, whether or not the petitioner could have raised the claim on direct appeal.” That is why the ordinary rules of procedural default do not bar these claims merely because appellate counsel did not raise them — and it is a mercy, because the trial record almost never contains the evidence a plea-advice claim needs.

A second §2255 motion is a different animal. If you have already filed one, a new plea-advice claim runs into the §2255(h) gate and requires authorization from the court of appeals before a district court may consider it. That gate is narrow and is explained at second or successive 2255. This is the strongest practical reason to get the first motion right: put every ineffectiveness theory in it, because the second one may never be heard.

Related failures of counsel have their own pages. If your lawyer never filed the notice of appeal you asked for, that is a distinct claim with a presumed-prejudice rule, at failure to file an appeal. If counsel represented a co-defendant or had another divided loyalty, see conflict of interest in a criminal case. The broader catalog of attorney failures courts have accepted is at ineffective assistance of counsel.

Applied Insight — Elizabeth Franklin-Best, Esq., Principal Attorney: The question we get most often is whether a claim is “strong enough.” That is not really answerable in the abstract, and anyone who answers it in a first phone call is guessing. What is answerable is narrower and more useful: does a document exist that contradicts what the plea record implies you knew? A written offer with no transmittal. A Guidelines worksheet with the wrong criminal-history category. A letter home, written before sentencing, describing a number nobody ever mentioned in court. When one of those exists, the case has a spine and the legal argument can be built around it. When nothing like it exists, the honest conversation is about what else might be recoverable — and about how much time is left on the clock while we look.

Why work with Elizabeth Franklin-Best, P.C.

Elizabeth Franklin-Best, P.C. is a boutique federal criminal defense and appellate firm founded in 2019, running a national federal practice on an intentionally small caseload. Elizabeth Franklin-Best has more than 20 years of federal and appellate criminal defense experience, has handled more than 330 federal proceedings, over 100 of them appeals, and has sixteen published decisions as counsel of record. She is admitted to the U.S. Supreme Court and twelve of the thirteen U.S. Courts of Appeals — every regional circuit (1st–11th) plus the D.C. Circuit — and has been admitted pro hac vice in federal courts from Montana to Florida.

She is recognized by Best Lawyers in America 2026 for Appellate Practice and ranked by Chambers USA 2026 for Litigation: White-Collar Crime & Government Investigations, was named South Carolina Public Defender of the Year in 2010, and is the author of Reversing Your Criminal Conviction: Develop Your Winning Strategy (2023). She serves as Board Chair of Justice 360 and as Co-Vice Chair of the NACDL Amicus Curiae Committee for the Fourth Circuit.

Plea-advice cases are record cases, and the firm’s approach to them is the one this page describes: identify which prejudice test governs, find the contemporaneous documents that make the showing concrete, and build the motion around what a court can see rather than around what a client remembers. With offices in Columbia and Mount Pleasant, South Carolina, the firm represents clients nationwide — a §2255 motion is filed in the district that imposed the sentence, wherever that is.

Talk to a federal post-conviction attorney about your plea

If you believe you were badly advised about a plea, an offer, or a consequence of a conviction, two questions matter first: what documents exist that show what you were told, and how much time is left on your filing deadline. The second one is urgent, and it does not wait while you assemble the first.

Schedule a consultation directly, or call (843) 620-1100. The firm handles federal post-conviction matters in every district and circuit. If you are still deciding what kind of representation your case needs, start at federal 2255 motion lawyer, which covers what post-conviction counsel actually does and how these engagements work.

Frequently Asked Questions About Ineffective Assistance in a Plea

Reviewed for legal accuracy by Elizabeth Franklin-Best, Esq., Principal Attorney·September 2026