To further delineate the mechanism responsible for the differences in xanthine oxidase activity in male and female Sprague-Dawley rats, a sensitive and specific radioimmunoassay (RIA) was developed for the measurement of hepatic xanthine oxidase. The RIA could detect as little as 5 mg of liver enzyme. Specificity of the RIA was confirmed by 1) Ouchterlony double immuno-diffusion in which a single precipitin band exhibited xanthine oxidase activity, when crude liver homogenate and an enzyme-specific stain were used; 2) parallelism between purified 125I-labeled xanthine oxidase and serial dilutions of crude liver homogenate; 3) a linear correlation between xanthine oxidase activity and the level of enzyme protein; and 4) a single protein band coincident with purified xanthine oxidase, when an immunoprecipitate prepared from antisera and crude liver homogenate was analyzed on sodium dodecyl sulfate (SDS) polyacrylamide gels. Whether xanthine oxidase activity was assayed in the absence of nicotinamide adenine dinucleotide (NAD+) (oxidase form) or in the presence of NAD+ (dehydrogenase), male values were consistently higher, and both forms of the enzyme correlated significantly with each other. When purified to homogeneity, neither form of the enzyme was appreciably affected by 17 beta-estradiol or testosterone propionate. When the RIA was employed, levels of hepatic xanthine oxidase were significantly greater in male than in female rats. We concluded from these data that increased xanthine oxidase activity in the male corresponds to a greater quantitative complement of xanthine oxidase protein. Furthermore, lower xanthine oxidase activity in the female cannot be explained by immunologically cross-reactive material without enzyme activity nor by a direct sex-steroid enzyme interaction.