In experiments in which liver and testis freeze-stops were performed on pentobarbital-anaesthetized rats, ethanol (1.5 g/kg body wt.) reduced plasma testosterone concentration from 13.1 to 3.2 nmol/litre. 4-Methylpyrazole abolished the ethanol-induced hepatic and testicular increase in the lactate/pyruvate ratio, and the testicular acetaldehyde level, but did not diminish the reduction in plasma testosterone concentration. In testes, but not in liver, ethanol decreased the 3-hydroxybutyrate/acetoacetate ratio, and 4-methylpyrazole did not prevent this effect. In experiments in which freeze-stop was performed after cervical dislocation, ethanol decreased the testis testosterone concentration from 590 to 220 pmol per g wet wt. The effects of ethanol and 4-methylpyrazole on testis acetaldehyde, lactate/pyruvate and 3-hydroxybutyrate/acetoacetate ratios were the same as found during anaesthesia. The NAD+-dependent ethanol oxidation capacity in testis ranged from 0.1 to 0.2 mumol/min per g wet wt. and seemed to be inhibited by 4-methylpyrazole both in vivo and in vitro. In additional experiments, ethanol doses between 0.3 and 0.9 g/kg body wt. did not alter the plasma testosterone concentration in rats treated, or not treated, with cyanamide, which induced elevated acetaldehyde levels in blood and testes. The results suggest that ethanol-induced inhibition of testosterone biosynthesis was not caused by extratesticular redox increases, or by extra- or intra-testicular acetaldehyde per se. The inhibition is accompanied by changes in testicular ketone-body metabolism.