Selective inhibition of estrogen production with aromatase inhibitors has been found to be an effective strategy for breast cancer treatment. Most studies have focused on inhibitor screening and in vitro kinetic analysis of aromatase inhibition using placental microsomes. In order to determine the effects of different inhibitors on aromatase in the whole cell, we have utilized the human choriocarcinoma cell line, JEG-3 in culture to compare and study three classes of aromatase inhibitors, 4-hydroxyandrostenedione, fadrozole (CGS 16949A), and aminoglutethimide. Fadrozole is the most potent competitive inhibitor and aminoglutethimide is the least potent among the three. However, stimulation of aromatase activity was found to occur when JEG-3 cells were preincubated with aminoglutethimide. In contrast, 4-OHA and fadrozole caused sustained inhibition of aromatase activity in both JEG-3 cells and placental microsomes, which was not reversed even after the removal of the inhibitors. 4-OHA bound irreversibly to the active site of aromatase and caused inactivation of the enzyme which followed pseudo-first order kinetics. However, 4-OHA appears to be metabolized rapidly in JEG-3 cells. Sustained inhibition of aromatase induced by fadrozole occurs by a different mechanism. Although fadrozole bound tightly to aromatase at a site distinct from the steroid binding site, the inhibition of aromatase activity by fadrozole does not involve a reactive process. None of the inhibitors stimulated aromatase mRNA synthesis in JEG-3 cells during 8 h treatment. The stimulation of aromatase activity by AG appeared to be due to stabilization of aromatase protein. According to these results, 4-OHA and fadrozole would be expected to be more beneficial in the treatment of breast cancer patients than AG. The increase in aromatase activity by AG may counteract its therapeutic effect and might be partially responsible for relapse of breast cancer patients from this treatment.