To study the influence of tetanus antitoxins on the outcome of human tatanus, we analyzed data on 545 cases reported to the Center for Disease Control from 1965 through 1971. Patients treated with antitoxin had a significantly lower case-fatality ratio than untreated patients, and the effect of serotherapy was not modified significantly by the age or race of the subjects. Antitoxin of equine origin and human tetanus immune globulin (TIG) were equally effective. The data on the effect of different doses of TIG suggest that 500 units may be as effective as the currently recommended therapeutic dose of 3,000 to 10,000 units. The importance of possible bias introduced by unmeasured factors such as quality of supportive therapy could not be determined, but confounding by 11 recorded potentially confounding factors was minimal.