This report describes a radioimmunoassay method for measuring unconjugated plasma testosterone (T) and its application to the measurement of T in murine plasma. The parameters of assay reliability, sensitivity, and accuracy were established. Di-hydrotestosterone, the only steroid which showed significant cross-reaction with the testosterone antibody used, was not found in significant quantities in male mice of 19 inbred strains examined. This suggested that chromatographic separation of T from other plasma steroids was unnecessary for the measurement of T in males of this species. Biological levels of T in normal adult male mice of 19 inbred strains showed a range of 0.64 to 21.9 ng/ml plasma and a mean of 5.10 ng/ml. Castration reduced testosterone levels to about 4% of the normal level; castration plus adrenalectomy reduced testosterone to levels below the sensitivity of this method. These investigations demonstrated the validity and reliability of this method for measuring uncanjugated plasma T in male mice.