Research Resources



An exoneration occurs when a person who has been convicted of a crime is officially cleared based on new evidence of innocence. This can occur through:

  • a pardon based on actual innocence
  • an acquittal at retrial
  • a conviction being vacated and indictment dismissed

DNA Exoneration

A DNA exoneration occurs when a person who has been convicted of a crime is officially cleared based on post-conviction DNA testing (i.e., the DNA testing results were dispositive of actual innocence and central to vacating the conviction and/or dismissing the indictment).

Much of the Innocence Project’s work focuses on cases where DNA evidence (e.g., blood or other bodily fluids) is central to the case, which tend to be cases involving sexual assault or murder. DNA exonerations represent only a portion, about 15%, of all exonerations in the United States.


As of January 2020, the Innocence Project has documented over 375 DNA exonerations in the United States. Twenty-one of these exonerees had previously been sentenced to death.

The vast majority (97%) of these people were wrongfully convicted of committing sexual assault and/or murder. Although these individuals were innocent of these crimes, approximately 25% had confessed and 11% had pleaded guilty. These exonerees spent an average of 14 years in prison–10% of whom spent 25 years or more in prison for crimes they didn’t commit.

Additional information about these DNA exoneration cases is available on our fast facts page, and a detailed analysis of the first 325 DNA exoneration cases (1989-2014) was published in the Albany Law Review in 2016. While the numbers have increased since then, the patterns and percentages of the various factors we addressed in the published paper remain remarkably consistent.

Data Sources for Research

The National Registry of Exonerations (NRE) is a public database that records all exonerations in the United States since 1989, including cases in which DNA played a limited or no role in the exoneration. In January 2020, the NRE database contained more than 3,300 cases. Their website is well suited to sorting and subsetting cases, and it provides graphic displays of case variables, including crime details and factors contributing to wrongful conviction. The site also has reports on special topics. Spreadsheets of the data are available upon request.

Convicting the Innocent is a public database developed in conjunction with Brandon Garrett, the author of the book Convicting the Innocent, an analysis of the first 250 DNA exonerations in the U.S. This database is limited to DNA exonerations and is updated annually. In January 2020 it included 367 cases. It has search capabilities and includes trial transcript excerpts and other source documents, as well as citations to judicial opinions.

The Work of the Innocence Project’s Data Science and Research Department

The Innocence Project’s data science and research department collects and analyzes data and conducts comprehensive reviews of research on issues related to wrongful convictions.

The data science and research department routinely collects data on DNA exonerations nationwide. In 2016, it published a detailed analysis of common factors in the first 325 DNA exoneration cases (1989-2014) in the Albany Law Review. In 2017, the department published an analysis that takes a closer look at the role of forensic science problems in wrongful convictions in the West Virginia Law Review. Lastly, in 2019, it published an examination of demographic crime-related and sentencing factors associated with innocent defendants who plead guilty in the Federal Sentencing Reporter.

To address the frequently asked question, “How common are wrongful convictions?”, the data science and research department critically reviewed the latest research and found that the wrongful conviction rate in capital cases is about 4% according to the best available study to date. This study was published by Samuel Gross and colleagues in 2014. A 2018 study by Charles Loeffler and colleagues reported an overall wrongful conviction rate of about 6% in a general state prison population, with considerable conviction-specific variability (from less than 1% to over 10%). This study provides some support for the previous estimate and reinforces the need for more research focusing on specific crimes and circumstances of conviction.

The data science and research department has also examined the influence of cognitive biases at various stages of criminal investigations and prosecutions. An overview of this work was presented at the 2017 meeting of the American Society of Criminology, and a summary of research on cognitive bias in forensic science that included 29 studies in 14 different disciplines entitled Cognitive Bias Research in Forensic Science: A Systematic Review was published in 2019 in Forensic Science International.

Just Data: Advancing the Innocence Movement

Just Data is the Innocence Project’s annual virtual gathering dedicated to elevating social science, connecting diverse stakeholders, and promoting practical research to advance the innocence movement.

Learn more about our event here. 

Other Research and Resources

Eyewitness Identification

The 2014 National Academy of Sciences report, Identifying the Culprit: Assessing Eyewitness Identificationexamines the scientific understanding of visual perception and memory and the implications of this understanding for developing best practices for investigating crimes and presenting eyewitness evidence. In 2020 Law and Human Behavior published an updated review of the social science literature with new information and recommendations.

False Confessions

One of the lessons learned from examining more than 375 DNA exonerations is that, although it may seem unimaginable, innocent people confess to crimes they did not commit for various reasons. Social scientists and legal scholars have studied the ways in which human psychology and specific interrogation techniques influence risk of false confession. From these studies, they also provide insights and recommendations.

Forensic Science

There has been considerable progress in some areas of forensic science since the 2009 National Academy of Sciences report, Strengthening Forensic Science in the United States: A Path Forward. The Center for Statistics and Applications in Forensic Evidence, initially founded in 2015, is now a multi-institutional, interdisciplinary collaboration with more than 60 researchers in statistics, computer science, machine learning, engineering and law. To highlight the contributions of statisticians working in this area, the Innocence Project helped coordinate a special issue of Significance, a journal by the Royal Statistical Society and the American Statistical Association, including articles discussing the need for objective measurementsreliability and validity, and the meaning of a “match,” co-authored by Innocence Project Staff Attorney Dana Delger.

Guilty Pleas

The Innocence Project contributed two papers to the April/June 2019 issue of the Federal Sentencing Reporter focused on “the trial penalty.” In addition to the data science and research department’s analysis of false guilty plea cases, Chris Ochoa and Deputy Chief Communications Officer Carlita Salazar provide a first-hand look at the experience of someone who faced the pressures and consequences of plea deal offer for a crime he did not commit.

Wrongful Convictions

The Ohio Innocence Project and the Innocence Project, in collaboration with the Innocence Network and the International Association of Chiefs of Police, created a series of educational videos introducing various ways that human factors (e.g., eyewitness memory) can impact criminal investigations and lead to wrongful convictions.

In 2019, the Innocence Project put together a recommended reading list and media guide for those interested in learning more about wrongful convictions through memoir, fiction, journalism, film and podcasts.

Innocence Research, a website created by four researchers interested in wrongful convictions, provides a collection of scholarship, popular media, upcoming conferences and meetings and other useful resources for teachers, policymakers, researchers and the general public.

Research Needs

The data science and research department periodically surveys Innocence Project staff to solicit input on unanswered questions that could have a direct impact on our work. We encourage future research to address these knowledge gaps. A summary of a 2017 staff focus group was presented at the 2018 American Psychology-Law Society conference.

If you are a researcher seeking to study exonerated people’s experiences or otherwise collect information from exonerated people through the Innocence Project or the Innocence Network, please submit an application to the Innocence Network Research Review Committee for approval.

Innocence Project Publications

Cooper GS, Meterko V. Cognitive bias research in forensic science: A systematic review. Forensic Sci Int. 2019;219:35-46.

Cooper GS, Meterko V, Gadtaula P. Innocents Who Plead Guilty: An Analysis of Patterns in DNA Exoneration Cases. Federal Sentencing Reporter. 2019;31:234-238. DOI: 10.1525/fsr.2019.31.4-5.234

Mejia R, Cuellar M, Delger D, Eddy B. What does a match mean? A framework for understanding forensic comparisons. Significance. 2019;16: 25-28.

Meterko V. Strengths and Limitations of Forensic Science: What DNA Exonerations Have Taught Us and Where to Go From Here. West Virginia Law Rev. 2016;119:639-649.

Ochoa C, Salazar C. How the Threat of the Trial Penalty Coerces the Innocent to Plead Guilty: A First-Hand Account of an Exoneree. Federal Sentencing Reporter. 2019;31:299-302. DOI: 10.1525/fsr.2019.31.4-5.299

West E, Meterko V. Innocence Project: DNA Exonerations, 1989-2014: Review of Data and Findings from the First 25 Years. Albany Law Rev. 2016;79:717-794.

We've helped free more than 240 innocent people from prison. Support our work to strengthen and advance the innocence movement.