THERE IS a widespread belief among the American public that people with mental illness pose a significant violence risk.^1,2 Moreover, the prevalence of this belief seems to have increased since the 1950s, when the issue was first systematically assessed.³ Because this public perception of dangerousness plays a central role in fostering stigma, its validity demands empirical scrutiny.^4,5 To date, nearly every modern study indicates that public fears are way out of proportion to the empirical reality. The magnitude of the violence risk associated with mental illness is comparable to that associated with age, educational attainment, and gender^6,7 and is limited to only some disorders and symptom constellations.^8,9 Furthermore, because serious mental illness is relatively rare and the excess risk modest, the contribution of mental illness to overall levels of violence in our society is minuscule.¹⁰