This prospective study evaluated changes in patients' health-risk behavior one year after preventive intervention by primary care physicians. The trial used a quasiexperimental design with 2,218 adults (1,409 study subjects and 809 controls) at five multispecialty, group practice sites in three regions of the United States. Pre- and postintervention surveys showed that the treated study patients with behavioral risks were more likely to report positive changes than were controls in regard to beginning regular exercise (P = .02), using auto seat belts, (P less than .001), losing weight (P = .05), decreasing alcohol intake (P = .01), and to performance of monthly breast self-examination by women (P less than .001). The smoking cessation rate was greater among the treated study group compared with the controls, although the change was not significantly different. Greater behavioral risk changes also occurred among the total study group (treated and untreated) in comparison with the control group. An additive index of these behavioral risks showed greater reduction among the treated and the total group of study patients compared with controls. These initial results suggest that clinical preventive services, which include risk factor education and counseling by primary care physicians, can improve short-term health-related behavior of patients.