Introduction In a meta-analysis of international studies, 17% of admitted patients in psychiatric hospitals had exhibited violent behavior toward others. Reported data from studies in Germany were considerably lower until recent years. However, studies examining only single hospitals, as well as the quality of the data itself, have raised questions as to the validity of these findings. Indeed, a debate currently exists as to whether there has, in fact, been an increase of violent incidents in German mental institutions. Methods In a group of 10 hospitals serving about half the population of the Federal State of Baden-Wuerttemberg with 11 million inhabitants, the Staff Observation Aggression Scale-Revised (SOAS-R) was introduced into patients' electronic charts as part of routine documentation. Data recording was strongly supported by staff councils and unions. A completed data set is now available for the year 2019. For one hospital, data are available since 2006. Due to some doubts with respect to fully covering self-directed aggression, we restricted the analysis to aggression toward others and toward objects. Results In 2019, 17,599 aggressive incidents were recorded in 64,367 admissions (1,660 staying forensic psychiatric inpatients included). 5,084 (7.90%) of the admitted cases showed aggressive behavior toward others. Variation between hospitals was low to modest (SD = 1.50). The mean SOAS-R score was 11.8 (SD between hospitals 1.20%). 23% of the incidents resulted in bodily harm. The percentage of patients showing violent behavior was highest among patients with organic disorders (ICD-10 F0) and lowest among patients with addictive or affective disorders (F1, F3, F4). Forensic psychiatry had the highest proportion of cases with aggressive behavior (20.54%), but the number of incidents per bed was lower than in general adult psychiatry and child and adolescent psychiatry (indicating a lower risk for staff). In the hospital with longer-term recordings available, an increase could be observed since 2010, with considerable variation between years. Conclusions This is the most robust estimate of the frequency of violent incidents in German psychiatric hospitals thus far. The incidence is about half of what has been reported internationally, probably due to sample selection bias in previous studies and a relatively high number of hospital beds in Germany. Available data suggest an increase of violent incidents over the last ten years; however, it is unclear to which extent this is due to increased reporting.