SUMMARY: 1. There is a very high correlation between the amounts of urinary androstenol and those of dehydroepiandrosterone, androsterone and aetiocholanolone in women with metastatic breast cancer. 2. In spite of this correlation, which implies a common precursor, tritiated dehydroepiandrosterone or testosterone are not metabolized to androstenol at the periphery (maximum conversion: 0·4% of injected dose) in amounts sufficient to account for the urinary androstenol. 3. As a means of evaluating 'androgen status' androstenol assays do not appear to be more useful than those of the 11-deoxy-17-oxosteroids.