Active surveillance for toxic shock syndrome (TSS) was established in 1986 in Los Angeles County and in the states of Missouri, New Jersey, Oklahoma, Tennessee, and Washington. Case reports were solicited through biweekly contact with all acute-care hospitals. One hundred sixteen definite and 63 probable cases were reported; 85% of the cases occurred in female patients and 15% in male patients. Among cases in females, 83 (55%) were menstrual; the mean age of the patients with menstrual cases was 23 years (range, 12-46 years). The overall incidence of TSS was 0.53/100,000. The cumulative incidence varied significantly by region, ranging from 1.23/100,000 in Oklahoma to 0.22/100,000 in New Jersey (P = .0001); the incidence in all other areas ranged from 0.39/100,000 to 0.70/100,000. The incidence of menstrual TSS was 1.05/100,000 women 15-44 years of age and peaked in women between the ages of 15 and 19 years at 1.52/100,000. The incidence was higher in whites than in nonwhites for both menstrual TSS (1.21/100,000 vs. 0.34/100,000, P = .002) and nonmenstrual TSS (0.30/100,000 vs. 0.14/100,000, P = .031). Our data indicate that TSS continues to be a cause of morbidity. Although there is underreporting in national passive surveillance, the proportion of menstrual cases reported through active surveillance was similar to that reported to the passive system in 1986.