Standards of Review on Federal Appeal

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

What is a standard of review on appeal?

The standard of review is how much deference the court of appeals owes the district judge on a given issue. Legal questions get de novo review with no deference. Factual findings get clear error. Discretionary calls get abuse of discretion. Issues nobody objected to get plain error, which almost always loses. Identify the standard first; it predicts the outcome better than the facts.

Key takeaways on standards of review

• Every issue in a federal appeal has a standard of review attached to it, and Fed. R. App. P. 28(a)(8)(B) requires the brief to state it “for each issue.”
• De novo is the friendliest: the appellate court decides the legal question fresh, giving the district court’s conclusion no weight.
• Clear error means a factual finding stands unless the reviewing court is “left with the definite and firm conviction that a mistake has been committed” — Anderson v. City of Bessemer City, 470 U.S. 564 (1985).
• Plain error has four requirements, and the fourth is discretionary — United States v. Olano, 507 U.S. 725 (1993). Most pages state two or three of them.
• On plain error the burden flips: “It is the defendant rather than the Government who bears the burden of persuasion with respect to prejudice” (Olano).
• Winning the issue is not winning the appeal. Preserved error still has to survive harmless-error review — beyond a reasonable doubt for constitutional error (Chapman v. California, 386 U.S. 18 (1967)).

The four-part plain-error test, stated in full

Most pages on this subject state two or three prongs. There are four, all required, and the fourth is discretionary — which is why an appellant can be right on the law and still lose.

The rule itself is short. Fed. R. Crim. P. 52(b): “A plain error that affects substantial rights may be considered even though it was not brought to the court’s attention.”

United States v. Olano, 507 U.S. 725 (1993), broke that sentence into the operative test, and Puckett v. United States, 556 U.S. 129 (2009), restated it in its modern form.

Prong	What it requires	The Court’s language	Who must show it
1. Error	A deviation from a legal rule that was forfeited, not intentionally waived	“[T]here must be an error or defect—some sort of ‘[d]eviation from a legal rule’—that has not been intentionally relinquished or abandoned, i.e., affirmatively waived, by the appellant” (Puckett)	Defendant
2. Plain	The error must be obvious, not debatable	“‘Plain’ is synonymous with ‘clear’ or, equivalently, ‘obvious.'” (Olano); “clear or obvious, rather than subject to reasonable dispute” (Puckett)	Defendant
3. Affects substantial rights	The error must have been prejudicial — it must have changed the result	“This is the same language employed in Rule 52(a), and in most cases it means that the error must have been prejudicial: It must have affected the outcome of the district court proceedings.” (Olano)	Defendant — “It is the defendant rather than the Government who bears the burden of persuasion with respect to prejudice”
4. Discretion	Even with 1–3 satisfied, the court of appeals may decline to correct the error	The court “should not exercise that discretion unless the error ‘seriously affect[s] the fairness, integrity or public reputation of judicial proceedings'” (Olano)	Defendant, as a matter of persuasion — there is no entitlement

Two clarifications Olano supplies that materially change the analysis.

Waiver is not forfeiture. Olano distinguishes them: “Deviation from a legal rule is ‘error’ unless the rule has been waived,” and “[m]ere forfeiture, as opposed to waiver, does not extinguish an ‘error’ under Rule 52(b).” Forfeiture is failing to make a timely assertion of a right; waiver is the intentional relinquishment of a known right. A forfeited claim gets plain-error review. A waived one gets nothing.

The remedy is not limited to the innocent. Olano states: “An error may ‘seriously affect the fairness, integrity or public reputation of judicial proceedings’ independent of the defendant’s innocence.”

Applied Insight — Elizabeth Franklin-Best, Esq., Principal Attorney: The fourth prong is where appellate briefs get lazy. Counsel proves error, proves it was plain, argues prejudice, and then stops — as if correction follows automatically. It does not. The fourth prong is a separate argument about the integrity of the proceeding, and it has to be written as one.

What plain error looks like applied: the broken plea agreement

Puckett v. United States, 556 U.S. 129 (2009), is worth reading because it shows the test doing real work on a sympathetic set of facts.

The question, as the Court framed it, was “whether a forfeited claim that the Government has violated the terms of a plea agreement is subject to the plain-error standard of review set forth in Rule 52(b) of the Federal Rules of Criminal Procedure.” The government had agreed to support a reduction for acceptance of responsibility and then opposed it at sentencing. Defense counsel did not object.

The Court held that plain-error review applied. A broken promise by the government is not exempt from Rule 52(b) merely because the breach is serious; the defendant still had to satisfy all four prongs, including prejudice and the discretionary fourth.

The lesson for a defendant is uncomfortable and worth stating plainly: even where the government does something clearly improper, failing to object at the time moves the claim into a framework the defendant usually loses. That is the entire argument for preserving error for appeal.

De novo review

De novo review — literally “anew” — means the court of appeals decides the question for itself, giving the district court’s answer no deference at all. It is the standard most likely to produce relief, and it applies to questions of law:

• Interpretation of a federal statute or the Constitution
• Whether an indictment states an offense
• Whether jury instructions correctly stated the law
• The legal conclusion at the end of a suppression ruling — whether particular facts amount to reasonable suspicion or probable cause
• Whether a prior conviction qualifies as a predicate under a sentencing enhancement
• The district court’s interpretation of the Sentencing Guidelines

Suppression rulings illustrate the split screen that runs through this whole subject. What the officer did and said — historical fact — is reviewed for clear error. Whether what the officer did amounted to a seizure, or whether the facts add up to probable cause, is a legal question reviewed de novo. The same ruling, two standards, depending on which part of it is being challenged.

For an appellant this means the framing of the issue is not cosmetic. An argument phrased as “the judge got the facts wrong” invites clear-error review. The same complaint phrased as “on these facts, the legal standard was misapplied” invites de novo review. Where the record honestly supports the second framing, that choice is worth more than any additional page of argument.

Clear error

Clear error governs findings of fact made by the district judge — after a suppression hearing, at sentencing, or in any bench proceeding.

The formulation federal courts apply comes from Anderson v. City of Bessemer City, 470 U.S. 564 (1985), quoting United States v. United States Gypsum Co., 333 U.S. 364, 395 (1948): “[a] finding is ‘clearly erroneous’ when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.”

Anderson is a civil case construing Fed. R. Civ. P. 52(a), not a criminal case, and it is worth being exact about that. Its value here is that it states the standard federal appellate courts apply to fact-finding generally, and it says out loud what makes the standard so hard to satisfy:

“This standard plainly does not entitle a reviewing court to reverse the finding of the trier of fact simply because it is convinced that it would have decided the case differently.”

And:

“If the district court’s account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently.”

That second sentence is the whole ballgame. “Plausible” is a low bar, and it is the bar. An appellant attacking a factual finding is not arguing that the judge was wrong; the appellant is arguing that no reasonable reading of the record supports what the judge found.

Credibility findings are the hardest of all, because the trial judge saw the witness and the panel did not.

Abuse of discretion

Abuse of discretion applies where the district court was making a judgment call the law commits to its discretion rather than deciding a question with one right answer:

• Evidentiary rulings under the Federal Rules of Evidence, including Rule 403 balancing
• Whether to grant a continuance
• Whether to sever counts or defendants
• Denial of a mistrial
• Limits on cross-examination, within constitutional bounds
• Most case-management decisions
• The ultimate sentencing decision

A ruling is not an abuse of discretion because a different judge would have ruled differently. It is an abuse when the decision rests on an error of law, on a clearly erroneous factual finding, or falls outside the range of choices a reasonable judge could make on that record.

The most common route to reversal under this standard is not arguing that the call was unwise. It is showing that the discretion was exercised on a mistaken legal premise — and the Supreme Court has said so in as few words as the point can be made. Koon v. United States, 518 U.S. 81 (1996): “A district court by definition abuses its discretion when it makes an error of law.” Koon adds that “[t]he abuse of discretion standard includes review to determine that the discretion was not guided by erroneous legal conclusions.”

That sentence is worth more to an appellant than it first appears. It means “abuse of discretion” is not a dead end whenever a legal premise is buried inside the discretionary call. If a district court excluded defense evidence because it misread a rule of evidence, the ruling is reviewed under a deferential label but decided on a legal question — and the deference largely evaporates. The appellate task is to locate the legal premise inside the discretionary decision and attack that, rather than arguing about the reasonableness of the outcome.

Mixed questions of law and fact

Many real appellate issues are neither purely legal nor purely factual, and the standard depends on which way the question leans.

Where the issue is essentially whether the district court correctly identified and applied a legal standard, review is de novo. Where the issue is essentially whether the historical facts satisfy a standard whose application is fact-bound and case-specific, review is deferential. Suppression rulings are the everyday example: historical facts for clear error, the legal significance of those facts de novo. Voluntariness of a confession, the existence of a seizure, and whether a defendant was “in custody” for Miranda purposes have each generated substantial appellate law about precisely where on that spectrum they sit.

For an appellant, the practical instruction is the same as elsewhere on this page: characterize the issue accurately, but characterize it. A mixed question briefed as a factual complaint will be reviewed as one.

Harmless error: winning the issue and still losing the appeal

This is the stage appellants are least prepared for. Even a preserved, well-argued, correctly decided issue does not produce relief if the error did not matter.

Fed. R. Crim. P. 52(a) is the source: “Any error, defect, irregularity, or variance that does not affect substantial rights must be disregarded.”

Two standards apply, depending on the nature of the error, and — critically — the burden is on the government.

	Constitutional error	Non-constitutional error
Governing case	Chapman v. California, 386 U.S. 18 (1967)	Kotteakos v. United States, 328 U.S. 750 (1946)
The standard	“[B]efore a federal constitutional error can be held harmless, the court must be able to declare a belief that it was harmless beyond a reasonable doubt”	Whether “the judgment was not substantially swayed by the error”
The Court’s language	Chapman, 386 U.S. 18	“But if one cannot say, with fair assurance, after pondering all that happened without stripping the erroneous action from the whole, that the judgment was not substantially swayed by the error, it is impossible to conclude that substantial rights were not affected” (Kotteakos)
Who bears the burden	The government	The government
What happens in doubt	Error is not harmless	“If so, or if one is left in grave doubt, the conviction cannot stand” (Kotteakos)

Kotteakos also disposes of the argument the government most often makes: that the remaining evidence was strong enough. “The inquiry cannot be merely whether there was enough to support the result, apart from the phase affected by the error. It is rather, even so, whether the error itself had substantial influence.”

The contrast with plain error is the single most useful thing on this page. On a preserved claim, the government must prove the error was harmless. On an unpreserved claim, Olano puts the burden on the defendant to prove it was prejudicial. Same error, same record, opposite burdens — and the only difference is whether trial counsel objected. See preserving error for appeal.

Structural error: the small category that is never harmless

A narrow class of constitutional errors is not subject to harmless-error analysis at all. These are defects in the framework within which the trial proceeds, rather than errors in the trial process itself, and because their consequences are unmeasurable they cannot be quantitatively assessed against the other evidence.

Arizona v. Fulminante, 499 U.S. 279 (1991), is the decision that organizes the distinction. It separates “trial error” — error occurring during the presentation of the case to the jury, which can be weighed against the rest of the evidence — from structural defects in the constitution of the trial mechanism, which cannot. Errors in the second category require reversal without any harmlessness inquiry.

The category is genuinely small — complete denial of counsel, a biased judge, exclusion of members of the defendant’s race from a grand jury, denial of self-representation, denial of a public trial, and defective reasonable-doubt instruction are the classic examples. Most trial errors are not structural, and an appellant who assumes otherwise will lose the harmlessness fight by default.

Sentencing: procedural error, then substantive reasonableness

Sentencing has its own two-step structure, and it is the most productive territory in federal appellate practice because the first step is largely mechanical.

Gall v. United States, 552 U.S. 38 (2007), sets it out:

“It must first ensure that the district court committed no significant procedural error, such as failing to calculate (or improperly calculating) the Guidelines range, treating the Guidelines as mandatory, failing to consider the §3553(a) factors, selecting a sentence based on clearly erroneous facts, or failing to adequately explain the chosen sentence—including an explanation for any deviation from the Guidelines range. Assuming that the district court’s sentencing decision is procedurally sound, the appellate court should then consider the substantive reasonableness of the sentence imposed under an abuse-of-discretion standard.”

Read that list carefully, because it is a checklist of appealable sentencing error:

Gall procedural error	What it looks like in a real case
Failing to calculate, or improperly calculating, the Guidelines range	A miscomputed offense level, criminal history category, loss amount, or drug quantity
Treating the Guidelines as mandatory	Language on the record suggesting the judge believed the range bound the court
Failing to consider the § 3553(a) factors	No engagement with the statutory factors counsel actually raised
Selecting a sentence based on clearly erroneous facts	Reliance on a PSR fact that was objected to and never resolved
Failing to adequately explain the chosen sentence	A thin or absent explanation — including for any variance from the range

The last row is the one defendants underuse. An inadequate explanation is itself reviewable procedural error, independent of whether the sentence was defensible.

Substantive reasonableness — the argument that the sentence was simply too long — is reviewed for abuse of discretion and is much harder to win. Procedural error is where sentencing appeals are actually won. Underlying legal questions about how a guideline should be interpreted are reviewed de novo; underlying factual findings, for clear error. See grounds for appeal.

Which standard applies to my issue?

This table maps common federal criminal appellate issues to the standard that governs them, and to what that realistically means.

Issue	Standard	Practical reading
Jury instruction misstated an element (objected to)	De novo	Strong posture — a legal question with no deference
Jury instruction error (not objected to)	Plain error	Weak — four prongs, defendant bears the prejudice burden
Suppression — historical facts	Clear error	Very difficult
Suppression — whether facts amount to probable cause	De novo	Strong posture
Sufficiency of the evidence	Jackson v. Virginia standard	Very difficult — evidence viewed in the light most favorable to the prosecution
Admission of evidence over objection	Abuse of discretion	Moderate; strongest where the ruling rested on a legal error
Admission of evidence, no objection	Plain error	Weak
Denial of a continuance	Abuse of discretion	Difficult
Guidelines calculation (objected to)	De novo on interpretation; clear error on facts	Strong posture — the most productive sentencing issue
Guidelines calculation (not objected to)	Plain error	Weak, though Olano prong 3 is sometimes met where the range was wrong
Adequacy of the sentencing explanation	Gall procedural review	Underused and often available
Substantive length of the sentence	Abuse of discretion	Difficult
Denial of counsel entirely	Structural	Reversal without harmlessness analysis
Ineffective assistance of counsel	Generally not reviewable on direct appeal	Belongs in a § 2255 motion — Massaro v. United States, 538 U.S. 500 (2003)

Sufficiency of the evidence sits in a category of its own

Sufficiency review is often listed as though it were an ordinary standard. It is not — it is the most deferential review in criminal law, and it is deferential to the jury, not to the judge.

Under Jackson v. Virginia, 443 U.S. 307 (1979), “[t]he relevant question is whether after viewing the evidence in the light most favorable to the prosecution, any rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt.” Every inference is drawn for the government. Credibility conflicts are resolved for the government. The panel is not asking what it would have concluded; it is asking whether a rational juror could have concluded what this jury did.

Two consequences follow. First, “the jury got it wrong” is close to unwinnable as an appellate argument, and a page that suggests otherwise is doing a reader a disservice. Second, where a sufficiency argument does succeed, the payoff is unusually large: reversal for evidentiary insufficiency bars retrial altogether under Burks v. United States, 437 U.S. 1 (1978), while reversal for trial error does not. See what happens after an appeal is granted.

The preservation mechanics matter here too. Sufficiency is ordinarily preserved by a motion for judgment of acquittal under Fed. R. Crim. P. 29, and an appellant who intends to argue that a finding is unsupported by the evidence must also, under Fed. R. App. P. 10(b)(2), “include in the record a transcript of all evidence relevant to that finding or conclusion.”

Applied Insight — Christopher Zoukis, JD, MBA, Managing Director: The question families ask is “how strong is our appeal?” The answerable version of that question is “which standard of review applies to each issue, and was it preserved?” Two issues that feel equally unfair can be worlds apart on those two variables, and a candid answer at the outset is more useful than an optimistic one.

One issue, several standards: how to break a ruling apart

The table above assigns one standard per issue for readability. Real appellate issues are rarely that tidy, and the most useful analytical habit in this area is to stop asking “what standard applies to this issue” and start asking “what standard applies to each component of this ruling.”

Almost every substantial district court ruling has three layers, and each layer carries its own standard:

Layer of the ruling	Standard	Example, from a suppression denial
What happened — historical facts	Clear error	What time the stop occurred; what the officer said; whether consent was given
What the law requires — the governing legal rule	De novo	What “reasonable suspicion” means; whether the exclusionary rule applies to this category of violation
Whether the facts satisfy the rule — application	De novo, or deferential, depending on how fact-bound the application is	Whether these particular facts amounted to reasonable suspicion

The same decomposition works on a sentencing ruling: what the defendant did is fact, how the guideline should be construed is law, and whether the conduct fits the guideline is application.

This matters because an appellant who attacks the wrong layer has chosen the wrong fight. Attacking the fact layer means arguing that no reasonable view of the record supports the finding — an extremely high bar. Attacking the law layer means arguing that the district court misunderstood the rule — a question the panel decides for itself. Whenever the record honestly supports the second framing, that is the argument to make.

Two cautions belong with this. First, circuits differ in how they articulate the standard for particular recurring issues, and how a given circuit phrases the review of, say, a Batson ruling or a Speedy Trial Act determination is a question to answer in that circuit’s own case law rather than from a general statement. Second, none of this framing changes what the record contains. Choosing the legal characterization is legitimate advocacy; describing the record inaccurately in order to reach it is not, and it is the fastest way to lose a panel’s confidence in the rest of the brief.

For the circuit-specific practice that governs how these arguments are received, see federal appeals court.

Why the standard is decided before the argument is written

Standards of review are not an academic frame placed over an argument after the fact. They determine what the argument has to prove, and the rules treat them that way: Fed. R. App. P. 28(a)(8)(B) requires the brief to include, “for each issue, a concise statement of the applicable standard of review.”

That requirement has a practical consequence for issue selection. If two potential issues are equally meritorious but one is preserved and the other is not, the preserved one is worth several times more — because the preserved issue puts the burden of proving harmlessness on the government, while the unpreserved one puts the burden of proving prejudice on the defendant.

This is also why raising every possible issue is counterproductive. Jones v. Barnes, 463 U.S. 745 (1983), holds that appellate counsel “does not have a constitutional duty to raise every nonfrivolous issue requested by the defendant,” and notes that “[e]xperienced advocates have emphasized the importance of winnowing out weaker arguments on appeal and focusing on one central issue if possible, or at most on a few key issues.” A brief leading with a plain-error argument and burying a de novo one has inverted its own case.

For how these arguments reach a panel and when argument is heard, see how long does an appeal take; for the narrow circumstances in which review comes before final judgment at all, see interlocutory appeal; and if appointed counsel has told you no issue exists, see anders brief.

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

Elizabeth Franklin-Best, P.C. is a boutique federal criminal appellate and post-conviction firm founded in 2019 with a national practice. 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 recognized by Best Lawyers in America 2026 for Appellate Practice and by Chambers USA 2026 for white-collar litigation.

The firm’s method on a new appeal starts with the record and the preservation question, because those two things fix the standard of review — and the standard of review determines which issues are worth briefing.

Talk to a federal appeals attorney

If you want an honest read on which issues in your record are actually reviewable, and under what standard, that is where a consultation starts.

Schedule a consultation · Phone: (843) 620-1100 · Columbia office: 3710 Landmark Dr #113, Columbia, SC 29204

Elizabeth Franklin-Best, P.C. represents clients in federal courts nationwide as a federal appeals lawyer. Consultations are confidential.

Related guides on appellate standards

• Federal appeals lawyer — the hub: how a federal criminal appeal works end to end
• How to appeal a federal conviction — where the standard gets stated, and when
• Grounds for appeal — the issues these standards are applied to
• Criminal appeals lawyer — counsel who can name the standard for each issue
• Appeal bond — where “substantial question” borrows from this analysis
• What happens after an appeal is granted — the relief each standard can produce

Frequently Asked Questions About Standards of Review

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