Timely investigation of children with acute flaccid paralysis, with collection of stool specimens for virus isolation, is the primary strategy used to detect wild poliovirus circulation. To determine the optimal timing of stool specimen collection, studies of wild and vaccine poliovirus excretion published between 1935 and 1995 were reviewed. Data were compiled from comparable studies, scatter plots of the data were created, and third-order regression lines were calculated. The data indicated that wild polioviruses were excreted by a majority of previously unvaccinated infants and young children for 3-4 weeks. The duration of viral shedding was reduced, however, among children who were previously vaccinated with inactivated poliovirus vaccine, who had preexisting antibodies to the infecting serotype, or who had previous intestinal infection with homologous poliovirus. These data suggest that the 14-day period after onset of paralysis is the interval with the highest probability of detecting wild poliovirus excretion in paralyzed children.