Ineffective Assistance of Counsel: Grounds, Standards & How to Raise It

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

What counts as ineffective assistance of counsel?

Ineffective assistance means counsel’s performance fell below an objective standard of reasonableness and that deficiency changed the outcome. Courts have accepted failures to investigate, to consult a needed expert, to object to a guideline error, and actual conflicts of interest. They reject anything a reasonable lawyer might have chosen as strategy — which is most of what movants raise.

Key takeaways on ineffective assistance of counsel

• The Sixth Amendment guarantees effective counsel, and § 2255(a) is where a federal prisoner raises it. The claim belongs in a collateral motion, not a direct appeal — Massaro v. United States, 538 U.S. 500 (2003).
• Courts start from a strong presumption of strategy. Strickland requires “a strong presumption that counsel’s conduct falls within the wide range of reasonable professional assistance.”
• “Doubly deferential” is a § 2254 standard, not a § 2255 one. On a federal prisoner’s motion the district court applies Strickland directly — still deferential, but one layer, not two.
• Filing the claim waives attorney-client privilege as to communications needed to prove or disprove it, and your trial lawyer will usually be ordered to respond by affidavit.
• The strongest claims are the ones the trial record cannot show — an uninvestigated witness, an unconsulted expert, an undisclosed conflict — and each requires evidence from outside that record.
• A failure to investigate is judged by the reasonableness of the investigation, not by whether the resulting choice can be described as tactical after the fact.

What counts as ineffective assistance of counsel?

The right comes from the Sixth Amendment: “In all criminal prosecutions, the accused shall enjoy the right … to have the Assistance of Counsel for his defence.” U.S. Const. amend. VI. The Supreme Court has long read that as a right to effective assistance, because, as the Court put it in United States v. Cronic, 466 U.S. 648 (1984), “[t]he right to the effective assistance of counsel is thus the right of the accused to require the prosecution’s case to survive the crucible of meaningful adversarial testing.”

For a federal prisoner, the vehicle is 28 U.S.C. § 2255(a) — a sentence “imposed in violation of the Constitution or laws of the United States.” A Sixth Amendment ineffectiveness claim fits the first of the statute’s four grounds; what the rest of the statute says is covered on what is a 2255 motion, and the wider catalog of claims that can be raised alongside it is on grounds for a 2255 motion.

The governing test comes from Strickland v. Washington, 466 U.S. 668 (1984). It has two components: the movant must show that “counsel’s representation fell below an objective standard of reasonableness,” and that the deficient performance prejudiced the defense, meaning “a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” A reasonable probability is “a probability sufficient to undermine confidence in the outcome.” How courts apply each of those prongs, element by element, is set out in full on what Strickland requires and how courts apply each of its two prongs.

What that test does not tell you is which conduct satisfies it. That is what the rest of this page is about, because a claim is not built out of a standard. It is built out of a specific thing a lawyer did or failed to do, evidence that it happened, and evidence about what would have happened otherwise.

The strong presumption that it was strategy

This is the obstacle most pages skip, and it is the one that decides more § 2255 ineffectiveness claims than the prejudice prong does.

Strickland instructs that “[a] court must indulge a strong presumption that counsel’s conduct falls within the wide range of reasonable professional assistance.” Harrington v. Richter, 562 U.S. 86 (2011), adds the calibration: “The question is whether an attorney’s representation amounted to incompetence under ‘prevailing professional norms,’ not whether it deviated from best practices or most common custom.” And Richter explains why courts hold that line — an ineffectiveness claim “can function as a way to escape rules of waiver and forfeiture and raise issues not presented at trial, and so the Strickland standard must be applied with scrupulous care, lest ‘intrusive post-trial inquiry’ threaten the integrity of the very adversary process the right to counsel is meant to serve.”

In practice, that presumption converts most complaints into non-claims. A lawyer who did not cross-examine a witness, did not call a family member, did not file a particular motion, or did not make a particular argument is presumed to have chosen not to. The movant’s burden is not to show that a better lawyer would have done it differently. It is to show that no competent lawyer, on what this lawyer knew at the time, could reasonably have made that choice.

There is one important limit, and it is the crack most successful claims go through. Strickland itself made the deference owed to a strategic judgment depend on the investigation behind it: “strategic choices made after thorough investigation of law and facts relevant to plausible options are virtually unchallengeable; and strategic choices made after less than complete investigation are reasonable precisely to the extent that reasonable professional judgments support the limitations on investigation.” Wiggins v. Smith, Warden, 539 U.S. 510 (2003), sharpened it into a rule a court can apply: “Strickland does not establish that a cursory investigation automatically justifies a tactical decision… Rather, a reviewing court must consider the reasonableness of the investigation said to support that strategy.”

That is the difference between a losing claim and a viable one. A claim that says this was a bad decision loses to the presumption. A claim that says this decision rested on an investigation that no reasonable lawyer would have stopped where this one stopped attacks the foundation the presumption sits on.

One layer of deference, not two

A point worth getting right, because a great deal of published writing gets it wrong. Richter‘s much-quoted line — that Strickland and § 2254(d) are both “highly deferential,” and “when the two apply in tandem, review is ‘doubly’ so” — describes federal habeas review of a state court’s ineffectiveness ruling under the deference provision AEDPA added in 1996. A federal prisoner’s § 2255 motion is not that. There is no state-court adjudication to defer to, so the district court applies Strickland itself.

That is not license to relax. Richter is explicit that “[e]ven under de novo review, the standard for judging counsel’s representation is a most deferential one,” because the attorney “observed the relevant proceedings, knew of materials outside the record, and interacted with the client, with opposing counsel, and with the judge.” The distinction matters for a different reason: when you read a case rejecting an ineffectiveness claim, check whether the court was applying Strickland or applying § 2254(d) to a state court’s application of Strickland. Cases in the second category are not measuring the same thing your motion will be measured against.

Applied Insight — Elizabeth Franklin-Best, Esq., Principal Attorney: The first thing we look for in evaluating one of these claims is not the mistake. It is the record of what counsel knew when the decision was made. A trial file that shows an investigator was retained, witnesses were contacted, and a choice was made is very hard to attack. A file showing that nothing was done, and a decision was made anyway, is a different case entirely — and it is Wiggins, not Strickland‘s presumption, that governs it. That is why we obtain trial counsel’s file before we form a view about the claim, not after.

What kinds of attorney failure have courts actually accepted?

The table below is the working taxonomy: the categories of conduct that have produced relief in federal courts, the leading authority for each, what the movant had to prove, and the reason claims in that category usually fail. It is organized by conduct, because that is how a case actually presents.

Table 1. Categories of ineffective assistance, the leading federal authority, the required showing, and the usual reason for rejection.

Category of failure	Leading federal case	What the movant had to prove	Typical reason courts reject the claim
Failure to investigate the defense	Strickland v. Washington, 466 U.S. 668 (1984); Wiggins v. Smith, Warden, 539 U.S. 510 (2003)	That the limits counsel placed on investigation were not supported by reasonable professional judgment, and that a reasonable investigation would have produced something that changed the outcome	The movant does not say what the investigation would have found. A court cannot weigh an unspecified discovery against the trial evidence
Failure to review material the prosecution will use	Rompilla v. Beard, 545 U.S. 374 (2005)	That counsel knew the Government intended to rely on a specific file or record and did not obtain it, even though it was readily available	Counsel did review it, or the record does not show counsel was on notice of the Government’s intent
Failure to conduct any pretrial discovery / to litigate a suppression issue	Kimmelman v. Morrison, 477 U.S. 365 (1986)	That the omission rested on a mistaken view of the law rather than a choice, and — for a Fourth Amendment claim — that the underlying claim was meritorious and the verdict would probably have differed without the evidence	The underlying motion would have lost. Failing to make a meritless motion is not deficient performance
Failure to retain or consult a needed expert	Hinton v. Alabama, 571 U.S. 263 (2014)	That counsel’s failure came from ignorance of the governing law (there, a funding statute) rather than from a considered choice among experts	Expert selection is “a paradigmatic example” of strategy when made after a thorough investigation; courts will not compare the retained expert’s qualifications to a better one
Failure to call a witness	Day v. Quarterman, 566 F.3d 527 (5th Cir. 2009)	Name the witness; show the witness was available and would have testified; set out the content of the testimony; show it would have been favorable	The claim rests on the movant’s description of what an unnamed or unavailable witness would have said. Uncalled-witness claims are “not favored” because the content is “largely speculative”
Actual conflict of interest	Cuyler v. Sullivan, 446 U.S. 335 (1980); Mickens v. Taylor, 535 U.S. 162 (2002)	An actual conflict that adversely affected counsel’s performance — identified as a specific thing counsel did or did not do because of the conflict	Only a theoretical or potential divergence of interests is shown. Mickens: an “actual conflict” is “a conflict of interest that adversely affects counsel’s performance”
Failure to object at sentencing to a guideline error	Michael S. Johnson v. United States, 313 F.3d 815 (2d Cir. 2002); Glover v. United States, 531 U.S. 198 (2001)	That the guideline calculation was wrong under a rule counsel should have known, and that there is a reasonable probability of a lower sentence	The objection would have failed on the merits, or the record shows the judge would have imposed the same sentence regardless
Complete failure of adversarial testing	United States v. Cronic, 466 U.S. 648 (1984), as limited by Bell v. Cone, 535 U.S. 685 (2002)	That counsel “entirely fail[ed] to subject the prosecution’s case to meaningful adversarial testing” — a failure that is complete, not at specific points	The complaint identifies discrete omissions, which Bell v. Cone holds are governed by Strickland, with prejudice required
A structural error counsel failed to preserve	Weaver v. Massachusetts, 582 U.S. 286 (2017)	Either a reasonable probability of a different outcome, or that the violation was “so serious as to render his or her trial fundamentally unfair”	Prejudice is assumed to follow automatically from the error’s structural character. It does not, when the error reaches the court through an ineffectiveness claim

Case-by-case treatment is discussed below.

Three categories are handled on their own pages, because each has a distinct test. Advice about a plea is covered on guilty plea ineffective assistance; a lawyer who never filed a requested notice of appeal is covered on my lawyer never filed the appeal I asked for; and appellate counsel’s omission of an issue is covered on ineffective appellate counsel.

Failure to investigate — the largest category

Strickland states the duty in one sentence: “counsel has a duty to make reasonable investigations or to make a reasonable decision that makes particular investigations unnecessary.” The corollary matters as much — “a particular decision not to investigate must be directly assessed for reasonableness in all the circumstances, applying a heavy measure of deference to counsel’s judgments.”

Wiggins is the case that shows what “directly assessed” means. Counsel there limited the mitigation investigation and later called it strategy. The Court refused the label and looked at the investigation: “a court must consider not only the quantum of evidence already known to counsel, but also whether the known evidence would lead a reasonable attorney to investigate further.” Prevailing professional norms — “American Bar Association standards and the like” — are “guides to determining what is reasonable.”

Rompilla v. Beard, 545 U.S. 374 (2005), pushes into harder territory, because there the client and his family were unhelpful. The Court held that “even when a capital defendant’s family members and the defendant himself have suggested that no mitigating evidence is available, his lawyer is bound to make reasonable efforts to obtain and review material that counsel knows the prosecution will probably rely on as evidence of aggravation.” The file at issue was “a public document, readily available for the asking at the very courthouse where Rompilla was to be tried.” The generalizable rule: an uncooperative client does not excuse a failure to obtain records counsel knows the Government will use.

Kimmelman v. Morrison, 477 U.S. 365 (1986), is the non-capital counterpart and the more useful analogue for most federal cases. Counsel there conducted no pretrial discovery at all, based on mistaken beliefs about what the Government was obliged to turn over. The Court held that failure “unreasonable, that is, contrary to prevailing professional norms,” and observed that the justifications offered “betray a startling ignorance of the law — or a weak attempt to shift blame for inadequate preparation.” Kimmelman also holds that a Fourth Amendment claim may underlie a Sixth Amendment ineffectiveness claim — but adds a cost: where the principal allegation is a failure to litigate a suppression issue, the movant must also prove “that his Fourth Amendment claim is meritorious and that there is a reasonable probability that the verdict would have been different absent the excludable evidence.”

Wiggins, Rompilla, and Kimmelman are all § 2254 cases arising from state convictions; Wiggins and Rompilla were also reviewed through § 2254(d)’s additional filter, which AEDPA added in 1996, a decade after Kimmelman. The Strickland propositions they state apply directly on a federal § 2255 motion.

What separates a viable claim in this category from a dismissed one is almost never the omission itself — it is whether the motion states what the missing investigation would have produced, with declarations attached. The four-element uncalled-witness showing, the records and expert categories, and what the filing has to attach are set out at ineffective assistance of counsel failure to investigate.

Failure to consult or retain an expert

Expert failures split cleanly along one line: whether the omission came from a choice or from not knowing the law.

Hinton v. Alabama, 571 U.S. 263 (2014), sits on the losing side of that line for the lawyer. Counsel knew his firearms expert was inadequate but did not ask for more funding, because he mistakenly believed a statutory cap left him nothing more to request. The Court held that deficient: “[a]n attorney’s ignorance of a point of law that is fundamental to his case combined with his failure to perform basic research on that point is a quintessential example of unreasonable performance under Strickland.”

The Court was equally clear about the other side of the line. “The selection of an expert witness is a paradigmatic example of the type of ‘strategic choic[e]’ that, when made ‘after thorough investigation of [the] law and facts,’ is ‘virtually unchallengeable,'” and the Court declined to “launch federal courts into examination of the relative qualifications of experts hired and experts that might have been hired.”

Harrington v. Richter shows the same principle producing the opposite result. Counsel there did not consult a blood expert; the Court held a state court could reasonably conclude that was within professional norms, because “[i]n many instances cross-examination will be sufficient to expose defects in an expert’s presentation,” and counsel is entitled to balance limited resources against the return.

So the viable expert claim is not “a better expert existed.” It is “counsel did not know an expert was available, or did not know what the field could establish, and never found out.”

Conflict of interest

Conflicts get a different framework, and it is more favorable to the movant in one respect and narrower in others.

Cuyler v. Sullivan, 446 U.S. 335 (1980), holds that “a defendant who raised no objection at trial must demonstrate that an actual conflict of interest adversely affected his lawyer’s performance.” The advantage is on the back end: “a defendant who shows that a conflict of interest actually affected the adequacy of his representation need not demonstrate prejudice in order to obtain relief.” No reasonable-probability showing is required once adverse effect is established.

Mickens v. Taylor, 535 U.S. 162 (2002), narrows the entrance in two ways worth knowing before you plead it. First, “actual conflict” is not a separate box to check: “[a]n ‘actual conflict,’ for Sixth Amendment purposes, is a conflict of interest that adversely affects counsel’s performance.” Second, the trial court’s failure to inquire changes nothing — a defendant must still show adverse effect “even if the trial court failed to inquire into a potential conflict about which it reasonably should have known.” And Mickens expressly reserved whether Cuyler applies to successive representation, calling it “an open question” so far as the Court’s jurisprudence is concerned. That reservation is why a claim built on a former-client conflict should be pleaded with an adverse-effect showing in the alternative.

Adverse effect means a specific, identifiable thing: a cross-examination not conducted, a defense not advanced, a cooperation opportunity not pursued, a plea not explored — traceable to the divided loyalty. What kinds of conflicts qualify and how each is proved is treated on conflict of interest counsel.

Failure to object — and why waiver and forfeiture are not the same thing

A failure to object is a distinct category because the objection’s absence changes the standard of review on appeal, which is often the real injury.

The vocabulary matters. In United States v. Olano, 507 U.S. 725 (1993), the Supreme Court drew the line: “Waiver is different from forfeiture. Whereas forfeiture is the failure to make the timely assertion of a right, waiver is the ‘intentional relinquishment or abandonment of a known right.'” A forfeited issue survives — reviewed on appeal for plain error under Fed. R. Crim. P. 52(b), which requires an error that is clear, that affected substantial rights, and that seriously affects the fairness or integrity of judicial proceedings. A waived issue is generally gone, because there is no “error” left to correct.

That distinction is what a failure-to-object claim is really about. Counsel who says nothing forfeits; counsel who affirmatively agrees may waive. Either way the appellate court never reviews the issue on the merits, and the § 2255 motion becomes the only remaining route.

Two limits keep this category honest. Failing to make a losing objection is not deficient performance — there is no duty to raise a meritless argument, and no prejudice can flow from one. And a structural error does not become automatically reversible just because it reaches the court through an ineffectiveness claim: Weaver v. Massachusetts, 582 U.S. 286 (2017), holds that where a public-trial violation “is neither preserved nor raised on direct review but is raised later via an ineffective-assistance-of-counsel claim, the defendant must demonstrate prejudice to secure a new trial.” The burden is “either a reasonable probability of a different outcome… or… that the particular public-trial violation was so serious as to render his or her trial fundamentally unfair.”

Where a claim was forfeited rather than raised, the ineffectiveness claim also interacts with the rules on procedural default — although a claim of ineffective assistance itself is generally exempt from that bar.

Ineffective assistance at sentencing

Sentencing is the most underused category in federal practice and often the most winnable, because the errors are documentary rather than testimonial.

Glover v. United States, 531 U.S. 198 (2001), removed the threshold that used to block these claims. The Seventh Circuit had required the sentencing increase to be “significant”; the Supreme Court rejected that, holding it error “to deny relief to persons attacking their sentence who might show deficient performance in counsel’s failure to object to an error of law affecting the calculation of a sentence because the sentence increase does not meet some baseline standard of prejudice.” The Court’s rule is flat: “any amount of actual jail time has Sixth Amendment significance.”

Michael S. Johnson v. United States, 313 F.3d 815 (2d Cir. 2002), applies it in a § 2255 posture and is worth reading for how the analysis actually runs. Counsel failed to object to a drug-quantity calculation squarely governed by a Guidelines application note. The Second Circuit held that “defense counsel’s failure to object to a sentencing calculation error that likely resulted in an increase in defendant’s period of incarceration constituted ineffective assistance of counsel,” rejected the district court’s reasoning that the same sentence was available under either offense level, and vacated for resentencing.

What makes these claims tractable is the evidence. The presentence report, the Guidelines manual in force, and the sentencing transcript are documents; the deficiency is usually visible on their face. That is a very different evidentiary problem from proving what an uncalled witness would have said.

There is also a cognizability reason to frame the claim this way. A freestanding argument that the district court misapplied the Guidelines is generally not reachable on collateral review at all — see claims not cognizable under 2255 — while the same error pleaded as counsel’s failure to object is a constitutional claim under § 2255(a). The framing is not a technicality; it is the difference between a claim the court can hear and one it cannot.

Sentencing-stage claims also include failures beyond arithmetic — mitigation never developed, a § 3553(a) argument never made, a departure or variance basis never identified. Those are investigation claims wearing sentencing clothes, and they are judged by Wiggins: not whether the choice was defensible, but whether the investigation behind it was. Each category, what it requires the movant to show, and how the months are computed are set out at ineffective assistance at sentencing.

When prejudice is presumed — and how narrow that really is

Cronic identified three circumstances in which prejudice need not be proved: a complete denial of counsel at a critical stage, because “a trial is unfair if the accused is denied counsel at a critical stage of his trial”; a case in which “counsel entirely fails to subject the prosecution’s case to meaningful adversarial testing,” which “makes the adversary process itself presumptively unreliable”; and circumstances where “the likelihood that any lawyer, even a fully competent one, could provide effective assistance is so small that a presumption of prejudice is appropriate without inquiry into the actual conduct of the trial.” Cronic also notes that ineffectiveness “is also presumed when counsel ‘actively represented conflicting interests,'” citing Cuyler.

Everything about how that framework is applied turns on Bell v. Cone, 535 U.S. 685 (2002). The Court there read Cronic‘s second category literally: “When we spoke in Cronic of the possibility of presuming prejudice based on an attorney’s failure to test the prosecutor’s case, we indicated that the attorney’s failure must be complete.” The claim in Bell was that counsel failed to oppose the prosecution “at specific points” — a failure to put on mitigation and a waiver of closing argument. That, the Court held, “is not of degree but of kind,” and is governed by Strickland, with prejudice required.

The practical consequence is that the Cronic categories almost never fit a real case. A lawyer who was present, filed motions, cross-examined witnesses, and argued at sentencing did not “entirely” fail, however badly he performed. Pleading Cronic in the alternative costs nothing; building the case on it usually costs the claim, because Strickland prejudice then never gets pleaded at all.

Applied Insight — Christopher Zoukis, JD, MBA, Managing Director: The version of this claim I hear most often from inside a facility is “my lawyer did nothing.” Almost always the docket shows otherwise — appearances, motions, a sentencing memorandum. That gap is not the client misremembering; it is that the client saw the parts of the representation that involved him, and the docket records the parts that did not. Pulling the full docket and the trial file first is what converts an impression into either a specific, provable omission or an honest answer that there is not a claim here.

Why this claim belongs in a § 2255 motion, not a direct appeal

Because the evidence you need does not exist yet.

Massaro v. United States, 538 U.S. 500 (2003), settled the question for federal defendants:

“We hold 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.”

The Court’s reasoning is the practical part. On direct appeal, “appellate counsel and the court must proceed on a trial record not developed precisely for the object of litigating or preserving the claim and thus often incomplete or inadequate for this purpose.” Where counsel acted, “the record may reflect the action taken by counsel but not the reasons for it,” leaving an appellate court with “no way of knowing whether a seemingly unusual or misguided action by counsel had a sound strategic motive or was taken because the counsel’s alternatives were even worse.” A district court on a § 2255 motion has tools an appellate court does not: it “may take testimony from witnesses for the defendant and the prosecution and from the counsel alleged to have rendered the deficient performance.”

Massaro also removes a trap: “failure to raise an ineffective-assistance-of-counsel claim on direct appeal does not bar the claim from being brought in a later, appropriate proceeding under § 2255.” The Court did note an exception in practice — “[t]here may be cases in which trial counsel’s ineffectiveness is so apparent from the record that appellate counsel will consider it advisable to raise the issue on direct appeal.” Which claim belongs in which proceeding is worked through on direct appeal vs 2255, and the mechanics of getting the motion filed are on how to file a 2255 motion. If your conviction is from a state court rather than a federal one, the parallel route is a petition for writ of habeas corpus.

What Massaro does not do is extend the calendar. The one-year period in § 2255(f) runs from the latest of four dates and is unaffected by the fact that the claim could not sensibly have been raised earlier, which is why the investigation an ineffectiveness claim requires has to be scheduled against that date rather than around it. Compute yours before you decide what the motion will contain, using the § 2255 deadline calculator.

What filing this claim does to attorney-client privilege

This is the consequence most people learn about after they file, and it belongs at the front of the decision rather than the back.

Raising ineffective assistance puts your communications with trial counsel at issue. Federal courts uniformly hold that doing so impliedly waives the privilege as to communications necessary to prove or disprove the claim. In practice the Government will ask the court to order your former lawyer to answer your allegations, usually by affidavit — and that affidavit becomes evidence weighed against your own account of the same conversations.

Two things follow that change how a motion should be drafted. The breadth of the waiver tracks how broadly you plead: a motion raising many generally framed ineffectiveness claims opens more of the file than one raising two specific claims. And whether the Government may use what it learns outside the § 2255 proceeding — at a retrial, for example — is a separate question that turns on your circuit and on whether a protective order was requested at the outset rather than after disclosure. Both axes, the governing tests, and how to limit exposure are set out in full on attorney-client privilege when you claim ineffective assistance.

The practical point for case evaluation: the disclosure is the price of the claim, and it is paid in advance. A strong, specifically pleaded claim usually justifies a correspondingly narrow waiver. A weak claim framed broadly can open a great deal of the file for very little return — and because § 2255(h) makes a first motion realistically the only one, that trade cannot be undone later.

What evidence actually supports the claim

An ineffectiveness claim is an evidentiary problem before it is a legal one. The standard is not hard to state; proving what a competent lawyer would have found is.

Trial counsel’s file. Notes, correspondence, investigator reports, retained-expert communications, and the absence of any of those. The file is what shows whether a choice was made after investigation or instead of it — the Wiggins question.

Declarations from uncalled witnesses. The Fifth Circuit’s formulation in Day v. Quarterman, 566 F.3d 527 (5th Cir. 2009), is the clearest statement of what these require: the movant “must name the witness, demonstrate that the witness was available to testify and would have done so, set out the content of the witness’s proposed testimony, and show that the testimony would have been favorable to a particular defense,” and that showing is required “for claims regarding uncalled lay and expert witnesses alike.” The court explained the reason such claims are disfavored: “the presentation of testimonial evidence is a matter of trial strategy and… allegations of what a witness would have stated are largely speculative.” A signed declaration answers all four elements at once; a paragraph describing what someone would have said answers none of them.

Expert declarations. Where the claim is that an expert was never consulted, an expert declaration stating what the field would have shown is what converts speculation into a fact the court can weigh.

Records the trial record does not contain. Medical, psychiatric, school, employment, and institutional records; documents from a co-defendant’s case; the prior-conviction file at issue in Rompilla. Where the material was in the Government’s hands and was never produced, the claim may be a suppression claim rather than — or in addition to — an ineffectiveness claim; that is a different test, set out on our Brady violation page.

The sentencing materials. The presentence report, the objections filed or not filed, the Guidelines manual in effect, and the sentencing transcript.

Your own declaration — necessary, and rarely sufficient on its own. A movant’s uncorroborated account contradicted by trial counsel’s affidavit and by the plea colloquy is the most common way a claim is resolved without a hearing.

Whether this evidence gets you a hearing turns on 28 U.S.C. § 2255(b): a hearing is required “[u]nless the motion and the files and records of the case conclusively show that the prisoner is entitled to no relief.” The objective at the pleading stage is therefore not to win — it is to make the existing record non-conclusive, which is exactly what a signed declaration from outside that record does. How that plays out is covered on 2255 evidentiary hearing, and the realistic odds and timelines are on 2255 grant rate statistics.

Applied Insight — Elizabeth Franklin-Best, Esq., Principal Attorney: The question we ask about every proposed ground is what a hostile reader could point to in the existing record to refute it in one sentence. If the answer is the plea colloquy, or trial counsel’s likely affidavit, or the sentencing transcript, then the ground needs corroboration from outside the record before it is filed — not after the Government answers. Building that corroboration is most of the work in a § 2255 case, and it has to happen while the § 2255(f) year is still running.

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, practicing nationwide on an intentionally small caseload. Elizabeth Franklin-Best 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 handled more than 330 federal proceedings, over 100 of them appeals, with sixteen published decisions as counsel of record. She is a 2026 Best Lawyers in America honoree for Appellate Practice and Chambers USA 2026 ranked for white-collar litigation, 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).

An ineffective-assistance case is built out of records and declarations, most of which do not exist until someone goes and gets them. Because the motion is filed in the sentencing court, the firm appears in whichever district entered the judgment. What that engagement involves at each stage is set out on the hub page for a federal 2255 motion lawyer, and the broader question of who handles a case after the direct appeal is finished is answered on post conviction attorney.

Schedule a federal post-conviction case review

If you believe your trial lawyer failed you, the two things worth establishing before anything else are whether the failure is documented anywhere outside your own memory, and when your one-year period under § 2255(f) began to run. Both can be answered from the criminal docket, the judgment, and the trial file.

Schedule a consultation or call (843) 620-1100. The firm represents clients in federal courts nationwide. For an overview of everything that remains available after a conviction becomes final, start with post conviction relief.

Frequently Asked Questions About Ineffective Assistance of Counsel

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