Poliomyelitis caused by wild poliovirus has been virtually nonexistent in the United States since 1980, and vaccine-associated paralytic poliomyelitis (VAPP) has emerged as the predominant form of the disease. We reviewed national surveillance data on poliomyelitis for 1960-1989 to assess the changing risks of wild-virus, vaccine-associated, and imported paralytic disease; we also sought to characterize the epidemiology of poliomyelitis for the period 1980-1989. The risk of VAPP has remained exceedingly low but stable since the mid-1960s, with approximately 1 case occurring per 2.5 million doses of oral poliovirus vaccine (OPV) distributed during 1980-1989. Since 1980 no indigenous cases of wild-virus disease, 80 cases of VAPP, and five cases of imported disease have been reported in the United States. Three distinct groups are at risk of vaccine-associated disease: recipients of OPV (usually infants receiving their first dose), persons in contact with OPV recipients (mostly unvaccinated or inadequately vaccinated adults), and immunologically abnormal individuals. Overall, 93% of cases in OPV recipients and 76% of vaccine-associated cases have been related to administration of the first or second dose of OPV. Our findings suggest that adoption of a sequential vaccination schedule (inactivated poliovirus vaccine followed by OPV) would be effective in decreasing the risk of VAPP while retaining the proven public health benefits of OPV.