This study describes the prevalence of adverse events and length of stay in forensic psychiatric patients with and without a restriction order. Detailed clinical and administrative information from medical records and written court decisions was gathered retrospectively from admission until discharge for a Swedish population-based, consecutive cohort of forensic psychiatric patients (n=125). The median length of stay for the whole cohort was 951days, but patients with a restriction order stayed in hospital almost five times as long as patients without. Restriction orders were related to convictions for violent crime, but not for any other differences in demographic or clinical variables. The majority of the patients (60%) were involved in adverse events (violence, threats, substance abuse, or absconding) at some time during their treatment. Patients with restriction orders were overrepresented in violent and threat events. Previous contact with child and adolescence psychiatric services, current violent index crime, psychotic disorders, a history of substance, and absconding during treatment predicted longer length of stay. Being a parent, high current Global Assessment of Functioning scores, and mood disorders were all significantly related to earlier discharge. In a stepwise Cox regression analysis current violent index crime and absconding remained risk factors for a longer hospital stay, while a diagnosis of mood disorder was significantly related to a shorter length of stay.