Evidence of B19 parvovirus infection was sought by in situ hybridisation with biotinylated probes in 65 tissue samples from 32 pregnancies (fetuses, products of conception and/or placentas). Twenty-seven samples were reactive and the results were confirmed by other methods for B19 virus detection in 22 cases. The other methods used were in situ hybridisation with 3H and 35S labelled probes; dot-blot hybridisation with biotin and 32P labelled probes; polymerase chain reaction assay; negative stain and thin section electron microscopy; and radioimmunoassay for B19 antigen. The five false positive results by in situ hybridisation with biotinylated probes were considered to be due to non-specific biotin capture and were more frequent with unfixed samples than with formalin fixed material. It was concluded that while biotinylated probes offered advantages over radioactive probes for detecting B19 DNA by in situ hybridisation, positive findings should be confirmed by other methods.