Direct fluorescent antibody assay (DFA) using monoclonal antibody and enzyme-linked immunosorbent assay (ELISA) for rapid detection of Respiratory Syncytial Virus (RSV) in nasopharyngeal secretions (NPS) were compared with conventional virus isolation and identification procedures in cell cultures. When 134 NPS were examined from infants and young children with acute respiratory tract infection, 42 (31%) were culture-positive for RSV and 31 of these were detected by the appearance of a typical cytopathic effect and identified by DFA either before or after its appearance, whereas 11 were identified as RSV-positive by DFA performed blindly on HEp-2 cell cultures 5 or 10 days after inoculation. DFA for RSV on NPS smears was positive in 33 (26%) cases, from seven of which RSV was not isolated. The same group of 134 NPS was tested for RSV detection by three commercial ELISA kits. The sensitivities of the three ELISA kits when compared with a combination of culture and DFA results, were comparable (53%, 51%, and 47% for Ortho, Kallested, and Abbott, respectively), whereas specificity was 100% for all three assays. In the group of 26 NPS detected as positive by both virus isolation and DFA, 20-22 (77-85% according to different kits) were found positive for RSV by the three ELISA assays. These data suggest that virus isolation is still critical for diagnosis of a fair number of cases of RSV infection. Of the two rapid techniques, DFA is a valuable complementary method, whereas ELISA still lacks sensitivity. However, both DFA and ELISA were able to detect RSV in 7 of 8 young patients with severe respiratory infection (pneumonia, bronchiolitis), thus permitting diagnosis of RSV infection at least two days before culturing.