Can you withdraw a federal guilty plea after sentencing?
No. 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.” Before sentencing, Rule 11(d)(2)(B) allows withdrawal on a “fair and just reason.” After it, your vehicle is an appeal or a §2255 motion.
On this page
Key takeaways on withdrawing a guilty plea
- Sentencing is the cutoff, not judgment entry and not the appeal deadline. Rule 11(e) is categorical: once sentence is imposed, there is no motion to withdraw a plea. There are only two remaining routes — direct appeal and collateral attack.
- There are three stages, not two. Before the court accepts the plea, Rule 11(d)(1) lets you withdraw “for any reason or no reason.” After acceptance but before sentencing, you need a fair and just reason under Rule 11(d)(2)(B) — or a rejected plea agreement under Rule 11(d)(2)(A). After sentencing, neither is available.
- Accepting your plea and accepting your plea agreement are different events. United States v. Hyde, 520 U.S. 670 (1997), holds that once the court has accepted the plea, the fair-and-just standard applies even though the plea agreement is still deferred.
- The circuits do not apply the same test. The Eleventh Circuit weighs four factors, the Fourth and First Circuits six, and the Sixth Circuit seven — and the Fourth Circuit frames the inquiry far more narrowly than the First or Ninth. Which circuit you are in changes the analysis.
- “I changed my mind” and “I got more time than I expected” are not fair and just reasons. The Sixth Circuit put it plainly in United States v. Bashara, 27 F.3d 1174 (6th Cir. 1994): the rule exists to undo “a hastily entered plea made with unsure heart and confused mind,” not to let a defendant “make a tactical decision to enter a plea, wait several weeks, and then obtain a withdrawal if he believes he made a bad choice in pleading guilty.”
- After sentencing, a bad-plea claim becomes a §2255 claim with a one-year clock. That is not a technicality. It is why Rule 11(e) matters: the rule that closes the withdrawal door is the same rule that routes you to 28 U.S.C. §2255.
Can you withdraw a guilty plea after sentencing?
No. In federal court the answer is categorical, and the rule says so in one sentence. Fed. R. Crim. P. 11(e) provides: “After 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.”
Read what that sentence actually does. It does not set a high bar for post-sentencing withdrawal. It does not say withdrawal requires manifest injustice, or extraordinary circumstances, or a compelling showing. It removes the motion entirely. There is no such thing as a post-sentencing motion to withdraw a guilty plea in federal court, and a district judge who receives one has no authority to grant it as filed. What the judge will typically do instead — if the filing says enough — is treat it as something else: a notice of appeal, or a motion under §2255. That recharacterization is a courtesy, not a right, and it carries consequences of its own, including on the successive-motion rules discussed at second or successive 2255.
This is where most readers of this page actually are. If you have already been sentenced — if a judge has said a number out loud and a judgment has issued — the question you typed into a search box has a different answer than the one you were looking for, and the honest version is more useful. Your plea can still be undone. It just cannot be undone by asking to withdraw it. The two doors Rule 11(e) names are the only two doors: a direct appeal, and a collateral attack under §2255. Which of the two your claim belongs in is a real decision with real consequences, and it is worked through at 2255 vs direct appeal.
One more thing about the timing. “Imposes sentence” means the oral pronouncement at the sentencing hearing — not entry of the written judgment, not the expiration of the appeal period, not the point at which the conviction becomes final for limitations purposes. The window for a withdrawal motion closes at the hearing. If your lawyer is going to move to withdraw, the motion has to be filed and decided before the judge pronounces sentence.
The three stages of a federal guilty plea
Most pages on this subject describe two stages — before sentencing and after. Federal practice has three, and the stage most often omitted is the one with the most generous standard. Rule 11(d) distinguishes between withdrawal before the court accepts the plea and withdrawal after acceptance but before sentence, and the difference between those two moments is the difference between an absolute right and a discretionary motion you will probably lose.
Rule 11(d)(1) is the forgotten provision: “A defendant may withdraw a plea of guilty or nolo contendere … before the court accepts the plea, for any reason or no reason.” That is not a standard. There is nothing to prove. If the judge has taken your plea under advisement and not yet accepted it, you can take it back because you woke up and felt differently about it.
The trap is that defendants and their families almost never know whether the plea has been accepted, because the plea and the plea agreement are separate objects and courts frequently accept one while deferring the other. That is exactly what United States v. Hyde, 520 U.S. 670 (1997), decided. The Court held that where a district court accepts the guilty plea but defers decision on the plea agreement pending the presentence report, the defendant “may not withdraw his plea unless he shows a ‘fair and just reason.'” As the Court put it, “[g]uilty pleas can be accepted while plea agreements are deferred, and the acceptance of the two can be separated in time.”
A citation-accuracy note that competitor pages get wrong. Hyde was decided in 1997, when the fair-and-just standard lived at Rule 32(e). The 2002 restyling of the criminal rules moved it into Rule 11(d)(2)(B) and moved the categorical after-sentencing bar into Rule 11(e). The substance did not change, but the citation did. When you read Hyde, or Moore, or Buckles, or Bashara, you are reading opinions written against Rule 32(d) or 32(e). They remain good law on the standard; they simply cite a rule number that no longer exists.
Table 1. The three withdrawal stages in federal court — what governs, what you file, and what actually happens
| Stage | Governing provision | What you must show | What you file | Realistic outcome |
|---|---|---|---|---|
| Plea entered, court has not yet accepted it | Fed. R. Crim. P. 11(d)(1) | Nothing. Withdrawal is available “for any reason or no reason” | An oral request at the hearing, or a written motion to withdraw the plea before acceptance | Granted as of right. The case returns to a not-guilty posture and heads back toward trial |
| Plea accepted, court rejects the plea agreement under Rule 11(c)(5) | Fed. R. Crim. P. 11(d)(2)(A); Rule 11(c)(5)(B) | Nothing beyond the rejection itself. The court must advise you personally and give you the opportunity to withdraw | Withdrawal on the record at the hearing where the court announces the rejection | Granted. Hyde confirms that on rejection the defendant “can ‘then’ withdraw his plea for any reason.” Rare in practice, because most agreements are Rule 11(c)(1)(B) recommendations the court is never bound by |
| Plea accepted, before sentence imposed | Fed. R. Crim. P. 11(d)(2)(B) | A “fair and just reason for requesting the withdrawal.” You bear the burden; review on appeal is for abuse of discretion | A written motion, usually supported by a declaration and often requiring new counsel | Discretionary and usually denied. A completed Rule 11 colloquy is powerful evidence against you, and delay is weighed heavily |
| After the court imposes sentence | Fed. R. Crim. P. 11(e) | Not available. Withdrawal is categorically barred | Nothing. The plea “may be set aside only on direct appeal or collateral attack” | The claim must be repackaged as a direct-appeal issue or a §2255 motion, each with its own deadline and its own bars |
Table 1. Federal plea-withdrawal stages under Fed. R. Crim. P. 11(d) and 11(e). Rule 11(c)(5) governs the court’s rejection of a plea agreement. This table states federal procedure only; state manifest-injustice standards are different and do not apply in a federal criminal case.
What is a “fair and just reason” to withdraw a plea?
Rule 11(d)(2)(B) supplies the phrase and nothing else. It does not define “fair and just,” and the Supreme Court has never comprehensively defined it either. What exists instead is a set of circuit-specific balancing tests, and they are not the same test in different words. They differ in how many factors they weigh, in what those factors are, and — most consequentially — in how narrowly the circuit frames the underlying question.
Two recent published decisions show the spread. In United States v. Milam (4th Cir. Aug. 13, 2025), the Fourth Circuit held that a “fair and just reason” is one that “essentially challenges the fairness of the Rule 11 proceeding,” and that a properly conducted colloquy “creates a strong presumption that the plea is final and binding,” leaving the defendant “a very limited basis to withdraw it.” Compare United States v. Arce-Ayala, 91 F.4th 28 (1st Cir. 2024), decided nineteen months earlier, which describes the identical standard as “liberal” and “permissive” — “but not toothless,” because “[l]iberal allowance is not to be confused with automatic allowance,” and a defendant has no “unfettered right to retract a guilty plea.”
Those are different starting points. The Fourth Circuit asks whether something went wrong at the plea hearing. The First Circuit asks a broader question about the totality of the circumstances, of which the plea hearing is one input. A motion drafted for one court can be legally irrelevant in the other.
Table 2. What each circuit weighs on a presentence motion to withdraw
| Circuit | Governing test | The factors | How the standard is framed |
|---|---|---|---|
| First | United States v. Arce-Ayala, 91 F.4th 28 (1st Cir. 2024) | Six, nonexclusive: (1) whether the plea was knowing, intelligent, voluntary and Rule 11–compliant; (2) the strength of the reason for withdrawal; (3) timing; (4) whether there is a serious claim of actual innocence; (5) whether a plea agreement was reached or breached; (6) prejudice to the government | “Liberal” and “permissive,” but “not toothless.” Courts look to “the totality of the relevant circumstances”; the six factors are not exclusive |
| Fourth | United States v. Moore, 931 F.2d 245 (4th Cir. 1991); applied in United States v. Milam (4th Cir. 2025) | Six: (1) credible evidence the plea was not knowing or voluntary; (2) credible assertion of legal innocence; (3) delay between plea and motion; (4) close assistance of competent counsel; (5) prejudice to the government; (6) inconvenience to the court and waste of judicial resources | Narrow. The reason must “essentially challenge[] the fairness of the Rule 11 proceeding.” A proper colloquy creates “a strong presumption that the plea is final and binding” |
| Sixth | United States v. Bashara, 27 F.3d 1174 (6th Cir. 1994) | Seven: (1) time elapsed between plea and motion; (2) presence or absence of a valid reason for not moving earlier; (3) whether the defendant asserted or maintained innocence; (4) the circumstances of the plea’s entry; (5) the defendant’s nature and background; (6) prior experience with the criminal justice system; (7) potential prejudice to the government | The rule’s aim is to undo “a hastily entered plea made with unsure heart and confused mind,” not to permit a tactical do-over |
| Ninth | United States v. Hernandez, 105 F.4th 1234 (9th Cir. 2024) | Categories rather than factors: “inadequate Rule 11 plea colloquies, newly discovered evidence, intervening circumstances, or any other reason for withdrawing the plea that did not exist when the defendant entered his plea.” For a “new reason,” a two-part test — subjective good faith plus objective reasonableness | Broad, and recently broadened further. The 2024 two-part test drew a dissent arguing it “is too far a stretch from Fed. R. Crim. P. 11(d)(2)(B)’s requirement” |
| Eleventh | United States v. Buckles, 843 F.2d 469 (11th Cir. 1988) | Four: (1) whether close assistance of counsel was available; (2) whether the plea was knowing and voluntary; (3) whether judicial resources would be conserved; (4) whether the government would be prejudiced | Totality of the circumstances. If the first two factors weigh against withdrawal, the court “need not give substantial attention” to the others |
Table 2. Circuit formulations of the Rule 11(d)(2)(B) “fair and just reason” standard. This is not a national rule. The circuits listed here have articulated tests; circuits not listed have their own formulations, and every one of these decisions is subject to intervening authority. Confirm the governing test in your circuit before relying on any row.
Two things are true in every circuit despite the divergence. First, the burden is on the defendant. Second, the standard of review is abuse of discretion, which means the district judge’s call will usually stand on appeal even if a different judge might have ruled the other way. Buckles states the appellate posture plainly: “[t]he district court may be reversed only if its decision is arbitrary or unreasonable.”
What is not a fair and just reason
The single most common reason a withdrawal motion fails is that the reason offered is one the defendant already knew about when he pleaded.
Changing your mind is not a reason. The Sixth Circuit’s formulation in United States v. Bashara, 27 F.3d 1174 (6th Cir. 1994), is the one most often quoted: the rule exists “to allow a hastily entered plea made with unsure heart and confused mind to be undone, not to allow a defendant ‘to make a tactical decision to enter a plea, wait several weeks, and then obtain a withdrawal if he believes he made a bad choice in pleading guilty.'”
A late claim of innocence, standing alone, is not a reason. United States v. Buckles, 843 F.2d 469 (11th Cir. 1988), states it in one line: “A mere declaration of innocence does not entitle a defendant to withdraw his guilty plea.” The reason is structural. You have already told a federal judge, under oath and on a transcript, that you did it. A bare contradiction of that statement does not carry weight against it. What can carry weight is a specific factual account — what the new evidence is, where it came from, why it was unavailable before — which is the difference the Supreme Court drew in Blackledge v. Allison, 431 U.S. 63 (1977), between “conclusory allegations unsupported by specifics,” which are “subject to summary dismissal,” and allegations that state “exactly what the terms of the promise were; when, where, and by whom the promise had been made.”
Getting more time than you expected is not a reason. This is the most painful one, because it is the most common real-world trigger. A defendant pleads on his lawyer’s estimate of a Guidelines range, the presentence report calculates it higher, and the sentence lands somewhere he never contemplated. The Ninth Circuit addressed exactly this pattern in United States v. Hernandez, 105 F.4th 1234 (9th Cir. 2024), reviewing its own precedent: where a defendant’s “true complaint” was that he had “underestimated the severity of the sentence that he faced” until he was “face-to-face with the full consequences of his conduct,” that was not a good-faith, fair and just reason for withdrawal, because “the defendant knew the downsides and the risks of his plea when he entered it.”
That last clause is the whole test in the Ninth Circuit’s framing. Hernandez held that to withdraw on a “new” reason, a defendant must show “(1) the ‘new’ reason is being offered in good-faith and that he subjectively did not know this reason for his plea withdrawal at the time of the plea, and (2) it was objectively reasonable to have not known about or anticipated this material new reason.” The disclosure the judge gives you at the plea colloquy — that the sentence is up to the court, that no one can promise you a number, that the Guidelines are advisory — exists precisely to make “I didn’t know I could get this much” objectively unreasonable later.
Delay is a reason to lose. Every circuit weighs it. In United States v. Spivak (6th Cir. July 22, 2026), the Sixth Circuit affirmed a denial where roughly three months separated the plea from the motion, noting that its cases “have treated delays of this length or shorter as excessive” — and it did so even though the district court accepted the defendant’s explanation that the delay came from a good-faith change of counsel rather than from a delay tactic. Buckles frames the relationship: “[t]he longer the delay between the entry of the plea and the motion to withdraw it, the more substantial the reasons must be as to why the defendant seeks withdrawal.”
Applied Insight — Christopher Zoukis, JD, MBA, Managing Director: When a family calls about withdrawing a plea, the first question is always the same, and it is not about the merits. It is has the judge sentenced him yet. If the answer is no, there is a live procedural motion with a real deadline, and the calendar is the emergency. If the answer is yes, we are having a completely different conversation about appeal deadlines and the §2255 clock, and the sooner that conversation starts the more of the record we can still get. The two situations look identical from the outside and share almost nothing procedurally.
What actually works: defects in the Rule 11 colloquy
If the reasons above fail, the reasons that succeed have a common shape. They point at something wrong with the plea proceeding itself, or with the advice that produced it — not at the defendant’s later regret.
Rule 11(b) is where to look, because it is the checklist the judge is required to complete. Before accepting a plea, the court “must address the defendant personally in open court” and both inform him of and determine that he understands a specific list of items. That list includes the right to plead not guilty and to persist in it; the right to a jury trial; the right to counsel at trial and at every other stage; the right to confront and cross-examine adverse witnesses, to be protected from compelled self-incrimination, to testify and present evidence, and to compel witness attendance; the waiver of those trial rights if the plea is accepted; the nature of each charge; “any maximum possible penalty, including imprisonment, fine, and term of supervised release”; “any mandatory minimum penalty”; any applicable forfeiture; the court’s authority to order restitution; the obligation to impose a special assessment; and the court’s obligation to calculate and consider the applicable Guidelines range.
Rule 11(b)(2) adds the voluntariness requirement separately: before accepting the plea the court “must address the defendant personally in open court and determine that the plea is voluntary and did not result from force, threats, or promises (other than promises in a plea agreement).”
The First Circuit distills those requirements into three “core concerns” of Rule 11 — “(1) a lack of coercion, (2) the defendant’s understanding of the charges against him, and (3) the defendant’s knowledge of the consequences of the guilty plea” — and holds that a failure to satisfy any one of them “requires that the guilty plea be set aside.” Arce-Ayala is itself an example of the third concern doing work: the First Circuit vacated where the district court’s own statements at the colloquy could have led a reasonable person to believe he would receive sentencing credit that a mandatory minimum made legally unavailable.
Other categories that succeed:
- A broken plea agreement. United States v. Moore, 931 F.2d 245 (4th Cir. 1991), restates the rule of Santobello v. New York: “when a plea rests in any significant degree on a promise or agreement of the prosecutor, so that it can be said to be part of the inducement or consideration, such promise must be fulfilled.” Where the government fails to adhere to the agreement, the remedy is either withdrawal of the plea or specific performance.
- A rejected plea agreement. Covered in the next section — this is the one situation where withdrawal after acceptance is available without any showing at all.
- Ineffective assistance in the plea advice itself. This is the largest category by volume and it has its own controlling test. Hill v. Lockhart, 474 U.S. 52 (1985), held that “the two-part Strickland v. Washington test applies to challenges to guilty pleas based on ineffective assistance of counsel,” with a plea-specific prejudice showing: “a reasonable probability that, but for counsel’s errors, he would not have pleaded guilty and would have insisted on going to trial.” The general two-prong framework from Strickland v. Washington, 466 U.S. 668 (1984), is set out at Strickland v. Washington, and the full plea-stage analysis — unconveyed offers, rejected offers, Guidelines misadvice — is at guilty plea ineffective assistance. A lawyer’s failure to warn about immigration consequences is its own claim, treated at Padilla v. Kentucky.
What happens if the judge rejects the plea agreement?
You get to withdraw, and you do not have to show anything. This is Rule 11(d)(2)(A), and it is the one route to post-acceptance withdrawal that carries no burden of persuasion.
The mechanics come from Rule 11(c)(5). If the court rejects a plea agreement that contains an agreed disposition — a specific sentence, a sentencing range, or the dismissal of charges the parties agreed would be dismissed — the court must advise the defendant personally in open court that it is not bound by the agreement, give the defendant the opportunity to withdraw the plea, and advise the defendant that if he persists in the guilty plea “the disposition of the case may be less favorable to the defendant than that contemplated by the plea agreement.” Hyde confirmed the consequence: “if the court rejects the agreement, the defendant can ‘then’ withdraw his plea for any reason and does not have to comply with … the ‘fair and just reason’ requirement.”
The practical caveat matters more than the rule. Most federal plea agreements are Rule 11(c)(1)(B) agreements, in which the government merely recommends a sentence or agrees that a particular Guidelines provision does or does not apply. A Rule 11(c)(1)(B) recommendation does not bind the court, so a judge who sentences above the recommendation has not “rejected” anything. Nothing triggers. The withdrawal right in Rule 11(d)(2)(A) attaches to Rule 11(c)(1)(A) and (C) agreements — charge bargains and binding sentence agreements — because those are the ones the court either takes or does not.
This is a distinction defendants very frequently misunderstand, and the misunderstanding is expensive: a defendant who believes his plea agreement guaranteed a number, and who is then sentenced above it, has usually signed a (c)(1)(B) agreement that guaranteed nothing about the sentence. That disappointment is not a rejected agreement, and it is not a fair and just reason. Whether the advice that produced the misunderstanding was constitutionally deficient is a separate question, and a real one.
Rule 11(h), harmless error, and the plain-error problem on appeal
Not every Rule 11 defect gets relief, and the rule says so in its own text. Rule 11(h) provides: “A variance from the requirements of this rule is harmless error if it does not affect substantial rights.”
That single sentence changed the landscape. It imports the harmless-error standard of Fed. R. Crim. P. 52(a) into Rule 11, so that a judge who omits an item from the colloquy has not automatically invalidated the plea. The question becomes whether the omission mattered.
On direct appeal, the burden depends on whether the error was preserved. If the defendant objected below, the government bears the burden of showing harmlessness. If the defendant did not — which is the ordinary case, because a defendant rarely interrupts his own plea colloquy to note a missing warning — the plain-error standard of Rule 52(b) applies and the burden flips.
United States v. Dominguez Benitez, 542 U.S. 74 (2004), supplies the showing. The Supreme Court held that “a defendant who seeks reversal of his conviction after a guilty plea, on the ground that the district court committed plain error under Rule 11, must show a reasonable probability that, but for the error, he would not have entered the plea.” The Court added two limits that do most of the work in practice:
- The reviewing court must be satisfied, “informed by the entire record, that the probability of a different result is ‘sufficient to undermine confidence in the outcome’ of the proceeding” — not just the plea proceeding, but the entire record, including the presentence report and the sentencing hearing.
- Structural error is not an escape hatch. “The omission of a single Rule 11 warning without more is not colorably structural.”
Two consequences follow for anyone reading this page. First, if you are still pre-sentencing, an unpreserved colloquy defect is far easier to raise now, by motion, than it will be later on appeal, because you will not be carrying the Dominguez Benitez burden. Second, if you have already been sentenced and are deciding between an appeal and a §2255 motion, the plain-error posture is a real factor in that decision — and a Rule 11 defect that was apparent on the face of the transcript usually belongs on direct appeal, because failing to raise it there can default it. That default rule comes from Bousley v. United States, 523 U.S. 614 (1998), which holds that “even the voluntariness and intelligence of a guilty plea can be attacked on collateral review only if first challenged on direct review.” The doctrine and its exceptions are covered at procedural default.
Applied Insight — Elizabeth Franklin-Best, Esq., Principal Attorney: The first document we want in a plea case is the change-of-plea transcript, and we want it before anyone theorizes about the claim. A colloquy defect either appears in that transcript or it does not, and reconstructing it from memory is how motions get filed on facts the record contradicts. If the transcript has not been prepared — and in a case that ended in a plea, it very often has not — ordering it is the first task, not a later one. Getting the transcript, the presentence report, and trial counsel’s file is its own project, and it is the one that determines whether a motion can be pleaded with facts instead of adjectives.
After sentencing, the plea becomes a §2255 question
This is the connection Rule 11(e) creates, and it is why a page about plea withdrawal ends up being a page about collateral review.
Rule 11(e) does not merely bar withdrawal. It names the two remaining vehicles — “direct appeal or collateral attack” — and for most claims about how a plea was actually induced, the second one is the only workable choice. The reason is evidentiary. A claim that your lawyer misdescribed the Guidelines range in a jail visiting room, or never told you an offer existed, or promised you a number he had no authority to promise, depends on facts that are not in the appellate record. An appeal is decided on the record. A §2255 motion can develop one.
The Supreme Court made that division explicit for ineffective-assistance claims in Massaro v. United States, 538 U.S. 500 (2003): “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.” So the Bousley default rule and the Massaro rule sit side by side: a Rule 11 defect visible on the transcript should generally go up on direct appeal or risk default, while an ineffectiveness claim about what counsel said off the record belongs in the §2255 motion and is not defaulted by omitting it from the appeal.
What §2255 is and what it can reach is set out at what is a 2255 motion; the mechanics of getting one on file correctly are at how to file a 2255 motion. Three constraints deserve mention here because they are the ones that catch plea cases specifically.
The one-year clock. Section 2255(f) gives a one-year limitation period running from the latest of four trigger dates. For most plea defendants the operative date is when the judgment of conviction becomes final. The computation is not intuitive and it is the single most common way a meritorious plea claim dies; it is worked through at 2255 statute of limitations, with the full post-conviction calendar at federal appeal deadline. If the deadline has already run, whether anything saves the motion is addressed at equitable tolling.
The custody requirement. Section 2255(a) reaches “[a] prisoner in custody under sentence of a court established by Act of Congress.” If your federal sentence — including any term of supervised release — has fully expired, §2255 is unavailable no matter how strong the plea claim is. Whether you are still in custody, and what the answer is if you are on supervised release or subject to a detainer, is worked through at the in-custody requirement. If custody has ended, the remaining vehicle is the one at coram nobis.
The record you will need. A §2255 motion has to plead facts, not conclusions — Rule 2(b)(2) of the Rules Governing Section 2255 Proceedings requires the motion to “state the facts supporting each ground,” and Rule 4(b) authorizes summary dismissal if it plainly appears from the motion and the record that the movant is not entitled to relief. For a plea claim that means the change-of-plea transcript, the sentencing transcript, the plea agreement, the presentence report, and counsel’s file. How to obtain each of them, and who pays, is at getting transcripts, PSR and your file. Whether you will get a hearing on the disputed facts is at 2255 evidentiary hearing.
The prison-mailbox timing rule, which can save a motion mailed before the deadline but docketed after it, is at prisoner mailbox rule. The broader AEDPA restrictions that shape all of this are at AEDPA, and if the conviction was in state rather than federal court the governing statute is different — see 28 U.S.C. § 2254.
Does an appeal waiver stop you from challenging the plea?
Often it does not, and the government’s broad framing of these waivers should not be accepted as settled.
Nearly every federal plea agreement now contains a waiver of the right to appeal and, increasingly, of the right to file a collateral attack. Those waivers are generally enforceable when knowing and voluntary. But there is a logical problem at the center of applying one to a claim that the plea itself was defective: the waiver is a term of the very agreement whose validity is in dispute. A waiver extracted by the same defective advice, or in the same defective colloquy, that produced the plea cannot supply its own validation.
Courts have long recognized some version of that point, and the scope of what survives a waiver — claims about the waiver’s own knowing and voluntary character, claims of ineffective assistance in negotiating the plea, sentences above the statutory maximum, and other categories that vary by circuit — is a substantial subject with its own live developments. It is covered in full at appeal waiver. The point to take from this page is narrower and practical: do not treat a waiver in your plea agreement as the end of the analysis. Whether it reaches a claim that the plea was involuntary or the product of ineffective assistance is a question to be litigated, not conceded.
How long do you have to withdraw a guilty plea?
There is no fixed number of days, and that is the answer most searches on this phrase are actually looking for. Rule 11(d) does not set a deadline in units of time. It sets one in units of events: the court’s acceptance of the plea, and the imposition of sentence.
- Before acceptance — unlimited, in the sense that no showing is required. The practical window is minutes to weeks, because most judges accept the plea at the change-of-plea hearing itself.
- Between acceptance and sentencing — however long that interval is in your case, which in federal practice is typically two to four months while the presentence report is prepared. Every day you wait counts against you. Buckles: “[t]he longer the delay between the entry of the plea and the motion to withdraw it, the more substantial the reasons must be.” Spivak: three months was too long in the Sixth Circuit even with a credited good-faith explanation.
- After sentence is imposed — zero. Rule 11(e) is absolute.
After sentencing, the deadlines that replace it are the ones that matter: 14 days to file a notice of appeal in a criminal case under Fed. R. App. P. 4(b)(1)(A), and one year under §2255(f), running from the latest of the four statutory trigger dates. Those two clocks do not run in parallel and are not interchangeable; a defendant who lets the appeal deadline pass has not extended the §2255 deadline, and a defendant who is still on direct appeal generally cannot get a §2255 motion decided in the meantime. Both are computed at 2255 statute of limitations.
Applied Insight — Elizabeth Franklin-Best, Esq., Principal Attorney: In a presentence withdrawal motion the thing we are building is not an argument, it is a record. Declarations from the people who were actually in the room. The written offer with its date. Emails and jail calls, if they exist, that show what the defendant was told and when. Judges deciding these motions are weighing a paper filing against a transcript in which the defendant said, under oath, that he understood everything and no one had promised him anything. The only way to move that weight is with contemporaneous documents. And if new counsel is needed because the claim is about prior counsel’s advice, that substitution has to be raised early — a motion filed by the lawyer whose advice is the subject of the motion has a structural problem before anyone reads it.
Why work with Elizabeth Franklin-Best, P.C.
Plea cases turn on records that are hard to get and harder to read: a change-of-plea transcript that was never ordered, a plea agreement whose (c)(1)(B) structure the defendant never understood, a presentence report that recalculated a range nobody predicted. Reconstructing what actually happened, and then matching it to the right vehicle and the right deadline, is the work.
Elizabeth Franklin-Best founded Elizabeth Franklin-Best, P.C. in 2019 and practices federal criminal appeals and post-conviction relief nationwide, on an intentionally small caseload. 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 district courts from Montana to Florida. Over more than 20 years she has handled more than 330 federal proceedings, over 100 of them appeals, with sixteen published decisions as counsel of record. She is recognized by Best Lawyers in America 2026 for Appellate Practice and by Chambers USA 2026 for Litigation: White-Collar Crime & Government Investigations, and is the author of Reversing Your Criminal Conviction: Develop Your Winning Strategy (2023).
Christopher Zoukis, JD, MBA, is the firm’s Managing Director and a nationally recognized federal prison expert. He is the author of the Federal Prison Handbook, Directory of Federal Prisons, Prison Education Guide, and College for Convicts, and serves as Chairman of the Board of the Human Rights Defense Center. He earned his J.D. at UC Davis School of Law after serving roughly a decade in the federal Bureau of Prisons, where he became an award-winning incarcerated journalist. Legal strategy and every filing in a federal case are handled by Elizabeth and the firm’s attorneys.
The firm serves clients nationwide from offices in Columbia and Mount Pleasant, South Carolina, and files in whichever district imposed the sentence. Who does this work and what they can still do after the appeal is over is described at post conviction attorney and federal 2255 motion lawyer.
Talk to a federal post-conviction attorney about your plea
If you have not been sentenced, the window is open and it is short. If you have been sentenced, the question is which vehicle and which clock — and the answer usually depends on documents that take weeks to obtain.
Schedule a consultation · Call: (843) 620-1100
Bring what you have: the plea agreement, the judgment, the docket sheet, and the date you were sentenced. If you do not have them, that is the first thing to fix, and getting transcripts, PSR and your file explains how.
Frequently Asked Questions About Withdrawing a Guilty Plea
Can you withdraw a guilty plea after sentencing?
No. Fed. R. Crim. P. 11(e) is categorical: “After 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.” There is no post-sentencing withdrawal motion in federal court at any standard. The plea can still be undone, but only through an appeal or a motion under 28 U.S.C. §2255.
How long do you have to withdraw a guilty plea?
There is no fixed number of days. Rule 11(d) measures the window in events, not time: unlimited before the court accepts the plea, discretionary between acceptance and sentencing, and unavailable once sentence is imposed. In the middle stage, delay counts against you — the Sixth Circuit affirmed a denial in United States v. Spivak (6th Cir. 2026) where about three months had passed, and United States v. Buckles, 843 F.2d 469 (11th Cir. 1988), holds that longer delay requires more substantial reasons.
What is a fair and just reason to withdraw a plea?
Rule 11(d)(2)(B) supplies the phrase without defining it, and the circuits apply different balancing tests — four factors in the Eleventh Circuit, six in the Fourth and First, seven in the Sixth. What generally works is a defect in the Rule 11 colloquy, a broken plea agreement, or ineffective advice that induced the plea. What generally fails is a changed mind, a bare claim of innocence, or a sentence higher than expected.
What happens if the judge rejects the plea agreement?
You may withdraw with no showing at all. Under Rule 11(c)(5) the court must advise you personally, give you the opportunity to withdraw, and warn that persisting may produce a worse outcome; Rule 11(d)(2)(A) then permits withdrawal. United States v. Hyde, 520 U.S. 670 (1997), confirms you can “then” withdraw “for any reason.” This applies to binding agreements, not to Rule 11(c)(1)(B) agreements where the government merely recommends a sentence.
Can I withdraw my plea if my lawyer misadvised me?
Before sentencing, bad advice can be a fair and just reason under Rule 11(d)(2)(B). After sentencing it becomes an ineffective-assistance claim in a §2255 motion. Hill v. Lockhart, 474 U.S. 52 (1985), supplies the test: deficient performance plus a reasonable probability that, but for counsel’s errors, you “would not have pleaded guilty and would have insisted on going to trial.” The full analysis is at guilty plea ineffective assistance.
Does an appeal waiver stop me from challenging the plea?
Not necessarily, and the government’s broad position on these waivers should not be accepted as settled. A waiver contained in the plea agreement cannot by itself validate a plea alleged to be involuntary or the product of ineffective assistance, because the waiver is a term of the same disputed agreement. What survives a waiver varies by circuit and is a live question — see appeal waiver.
This page is general legal information about federal criminal procedure and post-conviction law, not legal advice, and reading it does not create an attorney-client relationship. Federal plea-withdrawal standards differ by circuit, and every case turns on its own record and its own deadlines. Speak with a qualified federal criminal defense attorney about your specific circumstances.
Reviewed for legal accuracy by Elizabeth Franklin-Best, Esq., Principal Attorney·September 2026