Hospital data showed that there were 90,000 admissions for the treatment of injuries due to violent incidents between 1974 and 2002. The author recommends that hospital data on admissions for the treatment of injuries due to violence be included as another source in assessing trends in violence. There were more hospital admissions due to violence in the 1970s and 1990s and fewer in the 1980s. The hospital data did not show an increase in hospital admissions for victims of serious violence over this period. No increase was noted in either bone fractures or knife and gunshot wounds. Crime statistics, on the other hand, showed a continuous upward trend over this period. Hospital data showed the more stable trends in violence that were reported in victimization surveys and homicide statistics. This study used data from Sweden's Hospital Discharge Register, which records each admission, diagnosis and treatment, the sex and age of the patient, and the length of the hospital stay. The cause of an injury or illness is also noted. Various sources of error in the statistics regarding their relevance to trends in violence are discussed. 1 table, 6 figures, and 48 references