According to the National Institute of Justice, the total number of unprocessed rape kits in the country is unknown. Few states publish backlog data regularly, and among the ones that do, the asymmetrical reporting makes national comparisons difficult.
To help fill the data gaps, USAFacts gathered information from various state and city law enforcement agencies and crime labs nationwide. We asked every state for rape kit backlog data between 2018 and 2022, categorizing them into two groups: those awaiting testing at crime labs and those in possession of law enforcement. To standardize the data, we defined “backlog” as the number of kits left untested at a crime lab for more than 30 days and total number of kits unsubmitted by law enforcement at the end of each year. We sent these records requests in December 2022, allowing states six months to respond before publishing.
We received crime lab, law enforcement data, or both, from 30 states and Washington, DC. The formats of the information provided ranged from dozens of spreadsheets of kit-level data to one-page letters, and often required different types of analysis.
We combined this information with public reports to uncover some patterns behind the US rape kit backlog. For a deeper understanding of the state data and how each dataset’s limitations affected the numbers, see the analysis below.
Law enforcement agency backlog data
Time frame of data
States provided law enforcement data that covered varying time periods.
For instance, California and Montana’s 2018–2019 data only included kits collected within the respective reporting year. In contrast, Maryland’s data from 2018 to 2020 excluded kits backlogged from previous years that were subsequently tested by 2021 or 2022, while Tucson only included kits submitted in 2019 and later.
Some states were more comprehensive, presenting both current and legacy data. Massachusetts provided information on backlogged kits received in 2021 and 2022, along with those collected before 2018. North Carolina aggregated data from rape kits collected both before and after enacting the North Carolina Survivor Act in 2019, when all previously untested kits at law enforcement agencies were counted and required to be given to crime labs for testing.
Law enforcement reporting rates
Law enforcement reporting also varied. For instance, in Alaska, law enforcement participation dropped from 98% in 2019 to 75% in 2020. Louisiana had 100% participation from sheriffs’ offices during from 2018 to 2021, but police department engagement declined from 100% in 2018 to 67% in 2022. In New York, law enforcement participation fluctuated between 88% and 90% across the five years.
Numbers based on crime lab submission data
Data from Kentucky, Michigan, Washington, and West Virginia only reflected kits submitted to the crime lab by law enforcement. Therefore, the complete backlog of unsubmitted kits for each of those states remains unknown.
Calculation of days in law enforcement possession
The calculations of the number of days a kit spent in law enforcement custody varied due to differences in the provided kit-level data.
For Kentucky, Michigan, and Washington, we subtracted the date a kit was submitted to the lab from the offense date to determine the number of days that a kit remained in law enforcement possession. For West Virginia, we subtracted the date testing was requested for a kit from the offense date.
Several additional factors affected the backlog figures. For example, the Louisiana Commission on Law Enforcement questionnaire asked for a single backlog number but offered two definitions, increasing the likelihood of measurement inaccuracies.
Additionally, we obtained California’s law enforcement backlog by calculating the sum of kits where the suspect claimed consensual contact, kits with no specified reason for not being sent to the crime lab, and kits in transit to the lab.
Finally, for North Carolina, it was unclear if the counts of unsubmitted kit data we received accounted for those law enforcement held on to for such reasons as anonymity or victim’s consent declination.
Crime lab backlog data
Time frame of data
States also varied in the years of crime lab data they provided. Texas and West Virginia’s data, for instance, did not include kits collected before 2019 or 2018, respectively. New Mexico excluded kits collected before 2016.
On the other hand, California’s data only comprised kits that were collected within the respective year. Massachusetts included backlogged kits collected from 2000 to 2018, but excluded those received after 2018.
North Carolina’s dataset was more comprehensive, so we were able to aggregate kits collected both before and after the passing of the North Carolina Survivor Act in 2019. Note that North Carolina's crime lab backlog increased dramatically in 2019, when thousands of old untested kits held at law enforcement agencies were required to be given to crime labs for testing.
Crime lab reporting rates
Like law enforcement data, crime lab reporting rates vary significantly due to two factors: one, differences in laws on backlog data reporting; two, the rates at which crime labs comply with state law enforcement.
Alaska, Georgia, Idaho, Iowa, Louisiana, New York, North Dakota, Pennsylvania, South Dakota, Texas, and Vermont could confirm that their crime lab data represented their respective states.
However, we couldn’t verify whether California’s reports from 2018–2020 included data from all 58 counties, which is confirmed in California’s 2021 report. This potential omission of data from California’s largest counties may have affected California’s backlog data prior to 2021. (Los Angeles County processes rape kits locally, rather than sending them to the California Department of Justice, which compiles the annual reports.)
Maryland followed a similar pattern: only half of the forensic labs performing rape kit tests reported their backlog numbers to the Maryland Department of Justice.
Legislative requirements that require law enforcement to send backlog data to the state vary between states, as do law enforcement compliance with those laws.
Finally, New Mexico’s crime lab figures for 2018 and 2022 excluded kits processed in Albuquerque, the state’s most populous city.
The degree to which state crime labs relied on private labs for testing also differed. For instance, Kentucky’s numbers do not include kits that were outsourced to private labs. This was the case for 4,000 kits between 2015 and 2017, as well as for those currently being outsourced by the Louisville Metro Police. On the other hand, North Carolina provided rape kit data for its State Crime Lab and also publishes annual updates about rape kits from before 2019 that have been outsourced to vendor crime labs for testing; North Carolina's crime lab backlog numbers include all backlogged kits at both the state crime lab (about 3,500 in 2022) and outsourced vendor labs (about 5,500 in 2022).
The District of Columbia Department of Forensic Sciences lost its accreditation in April 2021, leading it to outsource all rape kit testing. Despite this, the department included data on these outsourced kits.
Several miscellaneous state-specific factors also influenced the data for public reports and data obtained through records requests alike. For California, we estimated crime lab backlogs using public reports, summing up kits undergoing analysis and kits that were left untested with an unspecified reason. In contrast, Maryland and Michigan’s crime lab data reflects only kits for which testing was completed, so the complete backlog of untested kits for both states remains unknown.
The District of Columbia forensic lab must receive permission from the courts to test before proceeding, so its backlog data excludes any kits that courts are processing.
Finally, the Iowa Division of Criminal Investigation provided estimates rather than exact counts for its backlog data.
If you or someone you know are a survivor of sexual assault, the Justice Department can help provide support and guidance.
In this article, we primarily highlight instances where data abnormalities might affect or limit how the data should be interpreted. When the data is fully accurate (to the best of our knowledge) and aligns with our standard definition of rape kit backlogs, we don’t usually mention it here.