Two methods of assaying alpha interferon (IFN-alpha) were compared during an experiment aimed at determining whether IFN-alpha crosses the human placenta. Human placentas, collected after delivery following a normal pregnancy to term, were catheterized on both sides: fetal and maternal. The IFN-alpha was introduced in known amounts in the maternal circulation and was assayed in the efferent fetal fluid. The following two detection methods were used: radioimmunoassay by competition with [125I]IFN-alpha and assay with a biological system in which IFN-alpha protected Madin-Darby bovine kidney cells from destruction by vesicular stomatitis virus. The results obtained by the two methods were in perfect agreement for the efferent fetal fluid samples. They showed the absence of placental transfer of IFN-alpha. The biological method was found to be more sensitive than radioimmunoassay for low IFN-alpha titers (< 10 IU/ml) but was less reproducible, probably owing to the use of twofold dilutions. The specificities of the two methods were similar and their practicalities were equivalent; the biological method, however, was less costly. The study illustrates the complementarity of the two methods, which were based on different principles. The agreement obtained between the two methods provides a clear confirmation of the experimental results.