Actual Innocence Standard in Post-Conviction Proceedings

To reopen a conviction under the actual innocence standard in post-conviction proceedings, a petitioner generally must present new reliable evidence and show that, in light of it, it is more likely than not that no reasonable juror would have found them guilty. That formulation comes from Schlup v. Delo, and in federal habeas practice it almost always operates as a “gateway” that lets a court reach constitutional claims otherwise blocked by procedural rules, not as a direct ticket to release.¹

What “Actual Innocence” Means in This Context

Actual innocence, sometimes called factual innocence, means you did not do the thing the government accused you of doing. That is a narrower concept than legal innocence, which can result from a procedural failure at trial such as an illegal search or a coerced confession even if the person physically committed the act. The distinction shapes which post-conviction route is available and how heavy the burden is.

Once a jury convicts, the presumption of innocence is gone. Courts treat the verdict as correct and approach factual innocence claims with intense skepticism, not out of indifference to wrongful convictions but because reopening settled cases based on new factual disputes cuts against the finality the system depends on. That tension between finality and accuracy is why the standard is deliberately steep.

The Schlup Gateway Standard

The gateway is the workhorse of actual innocence litigation. It applies when a petitioner has already defaulted their claims — by missing a deadline, filing a prior habeas petition, or failing to raise an issue in state court — and needs a way to have their constitutional arguments heard on the merits. To pass through, the petitioner must produce new reliable evidence and show that, more likely than not, no reasonable juror would have convicted in light of it.¹

Two features of the gateway are worth being precise about. First, clearing it does not free anyone. It opens the door so a court will look at the underlying constitutional violations the petitioner claims occurred at trial, such as ineffective counsel, suppressed evidence, or defective jury instructions. Without the gateway, those claims die on procedural grounds regardless of how serious they are. Second, the court is not confined to evidence that would be admissible at trial. It can weigh any relevant proof, including material that was wrongly excluded or simply unavailable earlier.

House v. Bell is the rare example of a petitioner clearing the gateway at the Supreme Court. Post-trial DNA testing showed that semen found on the victim’s clothing came from her husband, not from House, undercutting the prosecution’s sexual-motive theory. Combined with new witness testimony pointing to another suspect, the Court found this was “the rare case” where no reasonable juror viewing the full record would lack reasonable doubt.² Even then, House still had to proceed to a merits review of the constitutional errors he alleged.

Freestanding Innocence Claims

A freestanding claim is different. Here the petitioner is not trying to unlock some other constitutional violation. The argument is that the new evidence proves innocence outright, and that continued imprisonment or execution therefore violates the Constitution on its own.

The Supreme Court has never definitively held that freestanding innocence claims exist in federal habeas. In Herrera v. Collins, the Court assumed for the sake of argument that “a truly persuasive demonstration of actual innocence” after trial might make an execution unconstitutional, but declined to decide the question. The Court added that any such threshold would have to be “extraordinarily high” because pure-innocence reopenings would be deeply disruptive to the justice system.³

Lower federal courts generally acknowledge that freestanding claims might exist but almost never grant relief on that basis alone. If your only argument is “I didn’t do it” and there is no underlying constitutional error to attach the claim to, the federal habeas path is exceptionally narrow, and clemency from a governor or the president — the remedy the Herrera Court itself pointed to — is often the more realistic route.

Second or Successive Petitions: A Higher Bar

A petitioner who has already filed one federal habeas petition faces a tougher standard for any later one. Before a district court can even consider a second or successive petition, the federal court of appeals must authorize the filing. That court will only permit it when the claim rests on facts the petitioner could not have discovered earlier through due diligence and the new evidence, viewed with the rest of the record, would establish by clear and convincing evidence that no reasonable factfinder would have found the petitioner guilty.⁴ Clear and convincing is a heavier showing than Schlup’s more-likely-than-not, and most authorization requests are denied.

Getting Past the AEDPA Deadline and Exhaustion Rule

The Antiterrorism and Effective Death Penalty Act imposes a one-year deadline for filing a federal habeas petition. The clock generally starts when the conviction becomes final on direct appeal, though it can start later in certain situations, including when new facts surface that could not have been discovered earlier through reasonable diligence.⁴ Missing this deadline is one of the most common reasons habeas petitions never reach the merits.

In McQuiggin v. Perkins, the Supreme Court held that a credible actual innocence showing can override the one-year bar entirely, using the same Schlup standard. But the Court also warned that unjustifiable delay counts against the petitioner: how long the petitioner sat on the new evidence factors into whether the innocence showing is reliable.⁵ Sitting on exonerating evidence for years without explanation weakens the case considerably.

Before any federal habeas petition, the petitioner must also exhaust available state court remedies, giving state courts a fair opportunity to address the claims first.⁶ In practice, this means an innocence claim usually runs through years of state post-conviction litigation before a federal court sees it.

What Counts as New Evidence

The Schlup Court identified three broad categories of new evidence: exculpatory scientific evidence, trustworthy eyewitness accounts, and critical physical evidence not presented at trial.¹ “New” is doing real work in that phrase. Reinterpreting evidence the jury already saw rarely satisfies the standard. The evidence must be something the defense could not reasonably have obtained or presented during the original trial.

DNA and Discredited Forensics

DNA testing is the most powerful form of post-conviction evidence because it can definitively exclude someone as the source of biological material at a crime scene. Advances in testing methods now allow analysts to work with samples too degraded or too small for earlier examination. When the prosecution’s theory depended on the defendant’s physical presence, a DNA exclusion can gut the case.

Discredited forensic techniques open a different door. Older methods such as bite-mark comparison and certain types of hair analysis have been shown to lack scientific validity. When a conviction rested heavily on forensic testimony the scientific community no longer supports, courts may treat the discrediting of that methodology as new evidence. The National Registry of Exonerations has identified false or misleading forensic evidence as a contributing factor in roughly 29% of more than 3,600 exonerations recorded in the United States since 1989.

Recantations and Alternative-Perpetrator Evidence

Recantations can support a claim, but courts view them skeptically. A witness who changes their story years later faces obvious credibility problems, and courts look for a convincing explanation of why the original testimony was false and evidence that the new statement is voluntary. Perjury or false accusations have been identified as a factor in roughly 72% of known exonerations.

Evidence pointing to another perpetrator can be powerful but must go beyond speculation. The evidence needs to tie the other person to the actual commission of the offense in a concrete way, such as DNA at the scene matching that person or a corroborated confession.

Digital and Electronic Records

Cell phone location data, GPS records, and digital activity logs are increasingly used in post-conviction proceedings. They can establish an alibi, contradict a key witness’s timeline, or undermine claims about communications that allegedly occurred. Forensic analysts can recover call logs, text messages, and location data even from damaged or outdated devices. This kind of evidence is likely to matter more over time, particularly for convictions predating the smartphone era, where records once thought lost can now be pulled from old devices.

How Courts Actually Evaluate the Showing

For a gateway claim, the court conducts what amounts to a hypothetical retrial on paper. The judge combines the trial record with the new evidence and asks whether it is more likely than not that no reasonable juror would have convicted. The court does not have to find the petitioner definitely innocent. It only has to find the probability of innocence high enough to justify looking past procedural default and reaching the underlying constitutional claims.¹

Freestanding claims, to the extent they are recognized at all, demand a much higher showing. The Herrera Court described the threshold as “extraordinarily high,” and lower courts have generally read that as something beyond clear and convincing evidence, possibly approaching near-certainty.³ Almost no freestanding claim has succeeded in federal court.

Most petitions are initially reviewed on the papers alone. The judge reads the briefs, examines the new evidence, and decides whether the case warrants an evidentiary hearing. Cases tend to be won or lost at that hearing, where witnesses testify, experts present findings, and both sides cross-examine. A hearing signals the court takes the claim seriously but does not guarantee relief. When a court ultimately rules in the petitioner’s favor, the typical result is a vacated conviction and either a new trial or, in some cases, release when the prosecution declines to retry.

Statutory Post-Conviction DNA Testing

Separate from the innocence standard itself, federal law provides a statutory right to post-conviction DNA testing in federal cases through the Innocence Protection Act. To qualify, the petitioner must assert under penalty of perjury that they are actually innocent, identify a defense theory the testing could support, and show that testing might produce new evidence raising a reasonable probability they did not commit the offense.⁷ The evidence must still be in the government’s possession with an intact chain of custody, the testing must use scientifically sound methods, and where the petitioner was convicted after trial, identity must have been at issue.

The statute also covers situations where earlier DNA testing was performed but newer technology could produce substantially more probative results. A petitioner cannot use the provision if they knowingly declined to request DNA testing in an earlier motion.⁷ The Supreme Court held in District Attorney’s Office v. Osborne that there is no freestanding constitutional right to post-conviction DNA access, so state prisoners must rely on their state’s DNA testing statute, and the reach of those statutes varies widely.

Withheld Evidence as the Underlying Claim

Many wrongful convictions involve prosecutors who failed to turn over evidence favorable to the defense. Brady v. Maryland held that suppressing evidence favorable to the accused and material to guilt or punishment violates due process, regardless of the prosecution’s good or bad faith.⁸ When withheld evidence surfaces after conviction, it can function both as new evidence supporting the innocence showing and as the underlying constitutional claim that the gateway allows a court to reach on the merits. Official misconduct, which includes Brady violations along with coerced confessions, fabricated evidence, and witness intimidation, has been identified as a contributing factor in roughly 71% of known exonerations.

Pardon Versus Exoneration

One boundary is worth flagging because it changes what post-conviction relief actually gets you. A pardon is not the same as an exoneration. A pardon removes the legal penalties of a conviction but does not erase the conviction as a historical fact and does not imply factual innocence.⁹ A judicial exoneration is a finding that the conviction itself was wrong. Some compensation statutes require a formal finding of innocence rather than a simple reversal, so the legal basis on which a conviction is overturned can decide whether the exoneree qualifies for anything further.

1
Justia U.S. Supreme Court Center. Schlup v. Delo, 513 U.S. 298 (1995)
2
Library of Congress. House v. Bell, 547 U.S. 518 (2006)
3
Legal Information Institute. Herrera v. Collins, 506 U.S. 390 (1993)
4
Office of the Law Revision Counsel. 28 U.S. Code 2244 – Finality of Determination
5
Justia U.S. Supreme Court Center. McQuiggin v. Perkins, 569 U.S. 383 (2013)
6
Office of the Law Revision Counsel. 28 U.S. Code 2254 – State Custody; Remedies in Federal Courts
7
Office of the Law Revision Counsel. 18 U.S. Code 3600 – DNA Testing
8
Justia U.S. Supreme Court Center. Brady v. Maryland, 373 U.S. 83 (1963)
9
U.S. Department of Justice, Office of Legal Counsel. Whether a Presidential Pardon Expunges Judicial and Executive Branch Records of a Crime