Although toxic shock syndrome toxin—1 (TSST-1) has been proposed as the toxin responsible for toxic shock syndrome, its role in this disease has not been proved. To study this question, we examinedStaphylococcus aureusstrains isolated from normally sterile sites in patients with nonmenstrual toxic shock syndrome for the presence of TSST-1 production. Only 20 (62.5%) of 32 produced TSST-1, compared with 41 (93%) of 44 vaginal isolates from patients with menstrual toxic shock syndrome. Of strains of Saureusfrom patients with nonmenstrual toxic shock syndrome, TSST-1—negative isolates were more likely to be associated with a fatal outcome and to not be phage group I than TSST-1—positive isolates. Seven of the TSST-1—negative strains were evaluated in a rabbit subcutaneous chamber model of toxic shock syndrome. Fifteen (60%) of 25 rabbits developed a toxic shock syndrome—like illness and nine died. Clinical signs and histopathologic findings in the rabbits were similar to those seen in rabbits inoculated with TSST-1—positive Saureusisolates. These results suggest that other, as yet unrecognized, toxins play a role in toxic shock syndrome, and that TSST-1 production may not be essential to the pathogenesis of toxic shock syndrome. (JAMA1985;253:2538-2542)