A radioimmunoassay (RIA) was designed and compared with complement fixation and immunodiffusion tests for their relative ability to detect antibodies in sera of histoplasmosis patients. M antigen, purified from histoplasmin, was fixed to microtiter wells as the solid phase, and specific rabbit 125I-labeled anti-M globulin was the source of indicator antibodies. The optimal concentrations for the competitive-binding assay were 1.6 ng per well for M antigen and 650 ng per well for the 125I-labeled anti-M globulin. A panel of sera from 29 histoplasmosis patients and from patients with other mycoses was screened for RIA activity and in complement fixation and immunodiffusion tests that used histoplasmin and Histoplasma capsulatum yeast-form antigens. The sera of 22 histoplasmosis patients reacted in the RIA, 21 in the complement fixation, and 16 in the immunodiffusion tests. Sera of patients with other mycotic infections did not react in the RIA, with the exception of those of one blastomycosis patient and one candidiasis patient. The RIA could be modified to quantitate M antigen; as little as 125 pg could be detected. The evaluation of this panel of histoplasmosis patients' sera showed that the RIA was about equivalent in sensitivity to the complement fixation test. Some advantages of the RIA over the complement fixation test were that RIA was less prone to cross-reactions and gave better quantitation of low-titered sera. The RIA was a 1-day test, was not hindered by the anti-complementary activity of some sera, and could be modified to quantitate minute amounts of M antigen.