Keep Factually independent
Whether you agree or disagree with our analysis, these conversations matter for democracy. We don't take money from political groups - even a $5 donation helps us keep it that way.
Fact check: What is the estimated number of rapes in the US that go unreported each year?
Executive Summary
An exact annual count of rapes that go unreported in the United States cannot be pin-pointed with precision, but multiple authoritative sources converge on the conclusion that a substantial majority of sexual assaults are not reported to law enforcement each year, with common estimates ranging from roughly 50% up to about 70% or more of incidents going unreported. Surveys designed to measure victimization (the National Intimate Partner and Sexual Violence Survey and the National Crime Victimization Survey) and advocacy organizations (RAINN, NSVRC) provide differing methodologies and headline numbers, producing a range rather than a single definitive figure; this range implies hundreds of thousands to over a million sexual assaults annually that never enter criminal statistics [1] [2] [3] [4].
1. Why official crime counts understate the scale — Survey science shows the hidden majority
Survey-based victimization measures capture incidents that never reach police records, and that methodological gap explains why official arrest and police-report counts are much smaller than the true incidence of rape and sexual assault. The Bureau of Justice Statistics' National Crime Victimization Survey (NCVS) collects data directly from households about nonfatal personal crimes, including rape and sexual assault, and consistently finds higher incidence rates than police reports capture, though NCVS figures do not translate directly into a single “unreported” number year-to-year because of methodological limits and changes in survey design [2] [5]. Advocacy and research surveys such as the National Intimate Partner and Sexual Violence Survey (NISVS) provide prevalence estimates—NISVS found 2.3% of women and 0.3% of men reported rape in the prior 12 months in 2016/2017, producing millions of victims annually—while recognizing that sensitive topics are prone to undercounting even in anonymous surveys [1].
2. The frequently cited '70% unreported' headline — What it means and where it comes from
Multiple organizations and summaries cite that approximately 70% of sexual assaults go unreported, a figure that is often used in public discussion to highlight underreporting; sources such as the National Sexual Violence Resource Center and educational compilations repeat this proportion, and some public-facing summaries state that only about 20–30% of incidents are reported to police [4] [6]. That proportion is a useful shorthand for communicating scale, but it mixes different bases—some estimates refer to all sexual violence, others to rapes specifically, and the underlying data vary by year, question wording, age ranges, and whether the measure counts only forcible rape or a broader set of sexual victimizations; consequently the “70%” figure should be treated as an approximate indicator of widespread underreporting rather than a precise estimate [4] [7].
3. Counting the unreported: what the range implies in absolute numbers
Translating percentages into absolute counts produces wide-ranging estimates: NISVS’s prevalence estimates implied approximately 2.9 million women and 340,000 men experienced rape in the 12 months before the survey in the 2016/2017 measure, though NISVS cautions these are likely conservative [1]. RAINN’s public figure that about 433,000 people age 12 and older experience sexual violence each year is another commonly cited annual tally, and when combined with the assertion that around 70% are not reported, it implies hundreds of thousands of unreported incidents annually [3] [4]. The BJS criminal victimization reports (2023, 2024 summaries) provide baseline violent victimization rates but do not yield a single unreported-rape total; they do, however, demonstrate that the share of incidents captured by official statistics is substantially lower than survey-based estimates, reinforcing the scale of the unreported problem [5] [8].
4. Sources disagree on magnitude because they ask different questions — Methodology drives the spread
Differences across sources stem from variations in survey wording, the populations surveyed, definitions (rape vs. sexual assault), recall windows, and willingness of respondents to disclose traumatic events; these methodological choices cause the range of estimates. NISVS uses behaviorally specific questions to estimate prevalence, NCVS asks households about criminal victimizations and is designed to detect incidents not reported to police, and advocacy groups often compile or simplify statistics for public consumption, occasionally producing sharper-sounding claims like “less than one in three assaults are reported” or even lower percentages in some analyses [1] [2] [7]. Because no single study captures all incidents perfectly, multiple sources are needed to triangulate the scale of underreporting rather than to produce a single precise count [1] [2] [7].
5. What the numbers leave out and why that matters for policy and public understanding
Estimates focused on “unreported” incidents do not capture downstream effects like cases reported but not prosecuted, cases where victims withdraw cooperation, or demographic disparities in reporting by gender, age, race, and relationship to the perpetrator; these omissions mean policy responses based solely on headline underreporting figures risk overlooking justice-system attrition and survivor support needs. Advocacy organizations and research briefs note that underreporting is driven by stigma, fear of not being believed, and mistrust of authorities, and that improving measurement (better survey designs and victim-centered reporting options) and system responses (trauma-informed policing and prosecution) would alter both reporting rates and recorded totals—thus the best available evidence shows large numbers of unreported sexual assaults, but the precise annual tally depends on definitional and methodological choices that also shape policy solutions [7] [4] [8].