Eighty women who had onset of toxic-shock syndrome (TSS) between October 1, 1979, and September 19, 1980, and 160 age-and sex-matched neighborhood controls participated in this study to evaluate risk factors associated with TSS in women. Of the 80 women, 76 had onset of illness during their menstrual periods. The odds ratio for developing menses-associated TSS with any use of tampons compared with no use of tampons was 18.01 (P < 0.001). The odds ratios ranged from 5.29 to 27.5 for individual brand use compared with no use. When exclusive use of a particular tampon brand was compared with exclusive use of all other brands, Rely" (Procter and Gamble, Cincinnati, Ohio) was the only brand with a significantly increased odds ratio (2.49; P = 0.005). By multiple logistic regression analysis, the risk of TSS was more closely associated with tampon fluid capacity (absorbency) than with the use of all tampon brands.