A multiplex reverse transcription-PCR method was optimized to monitor the duration of excretion of Sabin poliovirus strains in stools of vaccinees following administration of the first dose of the trivalent oral vaccine. The assay detected approximately 1 50% tissue culture infective dose of each poliovirus serotype spiked into cell culture media. Although PCR inhibitors were frequently encountered in the stool specimens, a 1:20 dilution of the extracted RNA was sufficient to obtain a positive PCR result. Analysis of 195 stool specimens collected from 26 vaccinees showed that poliovirus types 1, 2, and 3 were identified more frequently by PCR than by tissue culture isolation. The percentages of specimens positive by PCR for poliovirus types 1, 2, and 3 were 67.2, 82.6, and 53.8, respectively. In contrast, the culture method identified types 1, 2, and 3 virus in 55.4, 64.1, and 27.7% of the samples, respectively. Poliovirus type 2 excretion was detected by PCR in practically all of the oral poliovirus vaccine recipients for 4 to 8 weeks following vaccination. In contrast, excretion of type 1 and 3 viruses was more variable, with a range of 1 to 8 weeks. Shedding of type 3 virus ceased in approximately 70% of vaccinees within a week after immunization. In addition to an enhanced sensitivity for the detection of poliovirus, this PCR method permits the direct characterization of virus in stool specimens without further passage in culture, which may select for genetic variants that may not accurately reflect the virus composition in the original specimen.