The Wisconsin and Centers for Disease Control (CDC) surveillance for toxic-shock syndrome (TSS) were compared; the distributions of CDC and Wisconsin cases in 1980 by calendar quarter of the initial onset were virtually identical. Among 131 cases in Wisconsin, 33 of 40 self-reported vs. 16 of 91 physician-reported cases were reported during the four-week period (September 18-October 15, 1980) following the CDC announcement of a higher relative risk of TSS among users of Rely tampons (Procter and Gamble, Cincinnati, Ohio) (P less than 0.001). Of cases with onsets during October 1, 1979-June 30, 1980, eight of 31 women with physician-reported cases vs. nine of 15 with self-reported cases used Rely tampons during their TSS-associated menstrual period (P less than 0.05). Multiple interviews of 22 patients with TSS suggested that patient recall of tampon brand may have been influenced by publicity about TSS. Passive as compared with partially active surveillance for TSS significantly favored self-reporting. In Wisconsin, the volume and source of TSS reporting have been greatly influenced by publicity, and selective reporting of TSS has been influenced by the content of news media reports of TSS.