Résumé : The new non-steroidal aromatase inhibitor vorozole (R83842) was administered orally at a single daily dose of 2.5 mg to 27 postmenopausal women with advanced breast cancer in a phase II trial as second-line endocrine treatment. The observed objective response rate of 30% and good tolerability of the drug confirm other recent reports. Endocrine determinations during 6 months of treatment demonstrated reduction of serum estrogens: estrone sulfate by more than 60%, estrone by 30-40%, but estradiol by only 10-20%. The ratios of serum androstenedione/estrone and testosterone/estradiol increased significantly, consistent with the inhibition of peripheral aromatase activity. Levels of progesterone, 17-alpha-hydroxyprogesterone, cortisol, dehydroepiandrosterone sulfate, androstenedione and aldosterone remained normal, indicating no interference with adrenocortical steroid synthesis. A general lack of correlation between the observed serum concentrations of vorozole and its effect on hormone serum levels or clinical response was found. This suggests that the determination of such serum levels gives a poor impression of the unambiguous anti-tumor activity of vorozole which may well have its main effect with the tumor tissue itself. The present results are in support of aromatase inhibition, but the possibility of an additional effect on the sulfation of estrogens merits further investigation.