Risk of suicide in people who have perpetrated specific forms of violent or sexual criminal offenses has not been quantified accurately or precisely. Also, gender comparisons have not been possible due to sparse data problems in the smaller studies that have been conducted to date. We therefore aimed to estimate these effects in the whole Danish population over a 26-year period. By completely interlinking national criminal, psychiatric, sociodemographic and cause-specific mortality registers, we conducted a nested case-control study of more than 27,000 adult suicides, during 1981-2006, and more than half a million age and gender-matched living controls. Elevated suicide risk was found in male sexual offenders. Risk was even higher among violent offenders, with greater effect sizes seen in females. It was markedly raised with serious violence, reaching a peak in relation to homicide or attempted homicide: male odds ratio (OR) 12.0, 95% confidence interval (CI) [8.3, 17.3]; female OR 30.9, CI [11.9, 80.6]. Following adjustment for psychiatric and social risk factors, relative risk in violent offenders was comparable to that seen among nonviolent offenders. These findings underline the importance of understanding why some people are violent toward themselves as well as other people, and why suicide risk is so much higher in people who have perpetrated serious acts of violence. They also indicate a clear need for developing effective multiagency interventions that effectively tackle both forms of destructive behavior.