In its 1982 report on toxic shock syndrome, the institute of Medicine, Washington, DC, identified population-based studies of the incidence of toxic shock syndrome over time based on hospital records as being a high priority for further research. We conducted such a study using records of hospitalizations in Northern California Kaiser-Permanente Medical Care Program facilities for the period from 1972 through 1983 among women and men aged 15 through 34 for illnesses considered possibly to be toxic shock syndrome. Of 6,688 hospitalizations reviewed, 54 were considered definite cases of toxic shock syndrome, and an additional nine were considered probable toxic shock syndrome. Four definite cases of toxic shock syndrome occurred in men, and 50 in women. The overall incidence of definite hospitalized toxic shock syndrome in men was 0.1 per 100,000 person-years, and in women 1.5 per 100,000 person-years. In women, an increase in the incidence of toxic shock syndrome was apparent by 1977; the rate peaked in 1980, decreased slightly in 1981 and 1982, and then almost doubled again in 1983. The temporal trend in the incidence of hospitalized toxic shock syndrome in women in the years 1977 through 1982 is consistent with the best available information on patterns of use of tampons containing higher-absorbency materials. The sharp increase in the incidence of hospitalized toxic shock syndrome in 1983 remains unexplained.