Human parasites are often distributed in metapopulations, which makes random sampling for genetic epidemiology difficult. The typical approach to sampling Schistosoma mansoni involves laboratory passage to obtain individual worms with small sample size and selection bias as a consequence. By contrast, the naturally pooled samples from egg output in stool or urine directly represent the genetic composition of current populations. To test whether pooled samples could be used to estimate population allele frequencies, DNA from individual cloned parasites was pooled and amplified by PCR for 7 microsatellites. By polyacrylamide gel analysis, the relative band intensities of the products from the major alleles in the pooled samples differed by 0–6% from the summed intensities of the individual clones (mean=2·1%±2·1% S.D.). The number of PCR cycles (25–40) did not influence the accuracy of the estimate. Varying the frequency of 1 allele in pooled samples from 32 to 69% likewise did not affect accuracy. Allele frequency estimates from aggregate samples such as eggs will be a better foundation for studies of parasite population dynamics as well as the basis for large-scale association studies of host and parasite characteristics.