The Static-99R is an actuarial risk assessment tool that scores adult men convicted of sex offenses on ten fixed historical factors, then sorts them into one of five risk levels tied to statistical recidivism rates. Parole boards, state sex offender registries, and civil commitment courts across the United States use the resulting score to decide how closely someone is supervised, how long they must register, and in the most serious cases whether they can be confined beyond their prison sentence. The score is mechanical by design: two trained evaluators working from the same records should land on the same number.
What the Ten Items Score
Every item on the Static-99R is historical. Nothing about treatment, remorse, behavior in custody, or time already served changes a single point. The items fall into three groups.
Age and relationships. The first item scores age at release, with younger men scoring higher. Someone released between 18 and 34 picks up a point; someone released at 60 or older loses three points. The second item asks whether the person has ever lived with a romantic partner for at least two years. No stable cohabitation adds a point.
Criminal history. Four items look at the record. One asks whether the current conviction involved non-sexual violence. Another asks the same about prior convictions. A third counts prior sex offenses, and this is the only item on the instrument that counts charges rather than convictions alone, meaning dismissed charges, arrests, and parole violations for sexual conduct can push the score up. A fourth asks whether the person has four or more prior sentencing dates for any offense.
Victim characteristics. The last four items each add a point for a “yes”: any convictions for non-contact sex offenses such as indecent exposure, any unrelated victims, any stranger victims, and any male victims.
Totals typically run from negative three to twelve.
How Scores Map to Risk Levels
The developers moved from four risk categories to five in 2016, and the five-level system is the one they currently endorse. Some older reports still reference the four-category version, so it is worth checking which scheme the evaluator used.
- Level I — Very Low Risk (scores of -3 to -2): predicted five-year sexual recidivism rate under 1.5%.
- Level II — Below Average Risk (scores of -1 to 0): roughly 2% to 3%.
- Level III — Average Risk (scores of 1 to 3): roughly 4% to 8%.
- Level IVa — Above Average Risk (scores of 4 to 5): roughly 11% to 15%.
- Level IVb — Well Above Average Risk (scores of 6 and above): starting at about 21% and climbing steeply with each additional point.
At the high end the numbers accelerate. A score of six corresponds to an estimated five-year recidivism rate of about 20.5%; a score of nine jumps to roughly 43.8%; a score of ten or higher corresponds to a predicted rate above 50%. Ten-year estimates run higher still. A score of five, for example, corresponds to roughly a 32% ten-year rate in high-risk samples.
Those percentages come from longitudinal studies that tracked thousands of convicted men over years and decades. They give courts a concrete statistical benchmark rather than a label like “dangerous.” They also, importantly, describe groups rather than individuals, a distinction covered further below.
Where the Score Shows Up in Legal Proceedings
The Static-99R appears at three main decision points, and the weight it carries differs sharply among them.
Parole and Probation Supervision
Parole and probation departments use the score to set the baseline for how closely someone is monitored. A higher score typically means more frequent check-ins, electronic monitoring, mandatory treatment, or geographic restrictions. Some states require the assessment by law for every eligible person before release. The score does not dictate conditions on its own; officers weigh it alongside other factors. But it anchors the starting point.
Sex Offender Registration
Several states fold Static-99R scores into their registration tier systems, using the risk level to help decide how long someone must register and whether their information appears on public notification websites. One boundary matters here: the federal Sex Offender Registration and Notification Act (SORNA) does not use risk assessment tools at all. SORNA assigns tiers based on the offense of conviction. States that layer risk-based classification on top of SORNA can produce situations where a person’s federal tier and state risk classification do not match.
Civil Commitment
The highest-stakes use is civil commitment, where the state seeks to confine someone in a treatment facility after their prison sentence ends. The U.S. Supreme Court upheld this type of indefinite civil detention in Kansas v. Hendricks (1997), holding that states may confine a person beyond their criminal sentence when a mental abnormality or personality disorder makes them likely to commit future sexual violence. The federal system has its own statute, 18 U.S.C. § 4248, which requires the government to prove by clear and convincing evidence that a person in federal custody is “sexually dangerous.”
Both state and federal civil commitment proceedings lean heavily on actuarial tools like the Static-99R to establish the risk component of that legal standard. Defense and prosecution experts routinely testify about the score, and disputes over how specific items should have been scored are common flashpoints at these hearings.
Who the Tool Applies To
The Static-99R was developed and validated exclusively for adult males aged 18 or older who have been charged with or convicted of at least one sexual offense against a child or non-consenting adult. The underlying research does not include women or juveniles, and applying the tool to them would produce unreliable results and invite legal challenge.
The person also has to be “at risk” in the community, meaning either already released or facing imminent release. Scoring someone who will spend the rest of their life incarcerated serves no predictive purpose, because the recidivism rates are calibrated to people who actually had the opportunity to reoffend. That is why the tool shows up at parole hearings and pre-release evaluations rather than at sentencing for the underlying offense.
A 2023 meta-analysis found that Indigenous and Black individuals scored significantly higher on the instrument than White individuals, though effect sizes were small. The tool’s ability to distinguish those who reoffend from those who do not was generally strong across groups but was lower for Indigenous and Hispanic individuals in some analyses. The findings do not render the tool unusable across racial groups, but they do mean scores for underrepresented populations warrant additional caution.
What the Score Cannot Do
Understanding the limits of the Static-99R is essential for anyone whose supervision, registration, or liberty depends on it.
Group rates, not individual predictions. The recidivism percentages tied to each score describe what historically happened within a group of people who all received that score. If the five-year rate for a score of six is 20.5%, that means roughly one in five people in that score group was reconvicted within five years. It does not mean any specific person with a score of six has a 20.5% chance of reoffending. Prosecutors sometimes present the group rate as if it were an individual probability. A prepared defense expert will flag this every time.
No account of current circumstances. Because every item is historical, the Static-99R cannot distinguish between someone who completed years of treatment and someone who refused it. A 60-year-old in good health and a 60-year-old with severe cognitive decline and physical disability receive identical scores if their records match. Evaluators are expected to note those circumstances separately, but the score itself does not reflect them.
Sample mismatch. The samples used to develop the Static-99R draw heavily from Canadian, British, Swedish, and other European populations. Criminal justice systems and reporting practices differ among those countries and differ further from the United States, which introduces uncertainty the score does not reflect on its face.
Challenging or Disputing a Score
There is no universal formal appeal process for a Static-99R score. The mechanism depends on the proceeding where the score is being used. In practice, disputes arise three ways.
Requesting a Rescore
Rescoring is appropriate when the original score came from a secondary source or when more complete information becomes available. If the criminal history records used in the original evaluation were incomplete or contained errors, an updated assessment with corrected records can produce a different score. What matters is that the factual inputs changed, not that someone simply disagrees with the result.
Identifying Coding Errors
Evaluators make mistakes despite the structured design. Common errors raised during expert testimony include scoring items on clinical impression rather than documented facts, counting the same criminal history twice (once through a scored item and again as an external aggravating factor), and inconsistent treatment of what qualifies as a “prior” versus the “index” offense. Forensic evaluators have been documented changing scores mid-testimony after hearing other experts, which undercuts the objectivity the tool is supposed to supply.
Arguing for a Departure
A departure is a written explanation that the actuarial score does not accurately reflect actual risk because of extraordinary circumstances. The classic example is a high-scoring individual who now has a severe physical disability and no realistic access to potential victims. The evaluator documents the calculated Static-99R score and then explains in writing why they believe it overestimates or underestimates risk. The score itself is not changed, because changing it voids the recidivism estimates attached to it. A departure is a narrative addition rather than a score modification.
In civil commitment and registration hearings, defense attorneys frequently retain their own forensic evaluators to independently score the instrument and challenge the state’s expert. Courts generally treat disagreements between dueling experts as credibility questions for the judge or jury rather than as evidence that the tool itself is invalid.
Checking the Evaluator’s Qualifications
The tool’s developers recommend an initial training of eight to twelve hours covering scoring rules, interpretation, and the empirical foundation. Fewer than eight hours is considered insufficient. Retraining is expected every two to five years, and anyone who has not been trained within the past ten years should not use the instrument. Trainees are expected to demonstrate competency through testing on case vignettes, not merely by attending a session.
Credential requirements vary by jurisdiction. Some states require a doctoral-level forensic psychology license. Others allow trained probation officers to score the tool for supervision purposes. The distinction matters in court: a score produced by a trained probation officer for internal classification purposes carries different evidentiary weight than one produced by a licensed forensic psychologist offering expert testimony. When reviewing a Static-99R report, it is reasonable to ask when the evaluator was last trained and whether they hold a competency-based certificate.