OBJECTIVES: An association between serum cholesterol concentration and violent behavior has been suggested, but has not been consistently demonstrated. This study was conducted in oreder to evaluate the relation between serum cholesterol concentration and violent behavior in psychiatric inpatients who had been admitted at Uijongbu St.Mary's Hospital, the Catholic University of Korea from January 1994 to June 1995. METHODS: We divided the 127 subjects into violent(35 subjects) and non-violent group(95 subjects). According to the percentile distribution of serum total cholesterol in healthy Korean adults, the subjects were classified into four subgroups:Group 1,subjects whose cholesterol concentrations were below 26% of percentile distribution, Group 2,from 26% to 50%, group 3, 51% to 75% and group 4, above 75%. In the four cholesterol subgroups, the authors examined the incidence of violent behaviors and suicidal attempts. Within the violent group, total aggression score of four cholesterol subgroups was compared. RESULTS: 1) Among the four cholesterol subgroups, there were no significant differences in the violent behavior. 2) within the violent groups, total violent score of the four cholesterol subgroups showed no significant difference. 3) Among the four cholesterol subgroups, there were no significant differences in suicidal attempt. CONCLUSION: When the serum cholesterol concentrations of psychiatric inpatient applied to the percentile distribution of serum total cholesterol in healthy Korean adult, no association was found between serum cholesterol concentration and violent behavior and suicide attempt.