Reported tetanus cases and tetanus deaths have declined substantially since widespread implementation of tetanus immunization. However, preventable morbidity and mortality continue to occur. During 1979-1984, 74-95 cases of tetanus were reported annually to the Centers for Disease Control (CDC) and 20-31 deaths were reported annually by the National Center for Health Statistics (NCHS). To evaluate further the health impact of tetanus, the authors assessed the completeness of national tetanus mortality data. They reviewed tetanus case report forms received at CDC from the states and NCHS multiple-cause-of-death tapes for 1979-1984. CDC reports recorded 129 deaths and NCHS mortality tapes 197 deaths. Year of death, state, age, and sex were used to match CDC and NCHS deaths, identifying 78 deaths reported to both sources. Using the methodology of Chandra Sekar and Deming, the authors estimated the actual number of tetanus deaths for 1979-1984 to be 326 (95% confidence interval 291-361). Based on this estimate, the completeness of reporting to CDC was 40%, to NCHS 60%, and to the combined systems 76%. To evaluate the reasons for underreporting, the authors contacted 14 states that had reported greater than or equal to 10 cases of tetanus to CDC during 1979-1984 to obtain death certificates for all tetanus deaths and additional information on all tetanus cases. Thirteen states submitted 108 death certificates for review. Coding and other systematic errors did not explain the low reporting efficiency. Failure to list tetanus as a cause of death on the death certificate was the primary reason for nonreporting of tetanus deaths to NCHS. These results suggest that NCHS tetanus mortality data may not be as complete as previously assumed and that tetanus mortality, and probably morbidity, are higher than previous reports have indicated.