Due to an unknown mechanism aminoglutethimide (AG) in urine samples is partly present in a form in which it is not detected. Though most evident if samples had been stored at -20 degrees C (about 60% trapped), it also occurred in fresh urine samples (about 20% trapped). Heating the urine sample for 15 min at 100 degrees C liberated all AG. This phenomenon has led to the underestimation of the excretion of unchanged AG in previous publications. This study shows that AG was excreted unchanged for 47.5 +/- 7.3% (mean +/- SD) in patients received 2 X 125 mg or 2 X 250 mg AG. Direct measurement of hydroxylaminoglutethimide (HxAG) was found difficult, due to its instability with respect to oxidation. We found azoxyG to be a stable oxidation product derived from HxAG. The concentration of azoxyG formed was proportional to the amount of HxAG converted. HxAG seemed subjected to the same phenomenon as AG. A conversion of HxAG into an unknown compound which seemed to occur very easily and which was even accelerated by heating, made it impossible to accurately measure the concentration of HxAG.