Putative specific and non-specie estrogen- 2 4 -hydroxylase activities which might affect the radio-enzymatic assay were characterized in terms of their requirements for NADPH and their substrate dependency. Using rat brain microsomes and partially purified rat liver COMT three main sources of estrogen- 2 4 -hydroxylase activity could be distinguished; a COMT-related component and NADPH-dependent and NADPH-independent microsomal components. The COMT-related activity required NADPH and showed about equal preferences for estrone and estradiol. The NADPH-dependent component was highly specific for estradiol, the relative activities observed with estrone and estriol being 7 and 1% of that observed with estradiol. The NADPH-independent component exhibited substrate saturation, was heat-labile and could not be inhibited by α-naphthoflavone or metyrapone. It showed a preference for estrone over estradiol, with estriol being a very poor substrate. These findings indicate that non-enzymatic factors contribute very little to product formation in the radio-enzymatic assay. The specificity of the major NADPH-dependent microsomal component towards estradiol suggests a stereo-specific requirement for the d-ring configuration of this estrogen. The use of no-cofactor blanks in the radio-enzymatic assay may be very important when different estrogens are compared as substrates for estrogen- 2 4 -hydroxylases.