Abstract. Plasma testosterone (T, nmol/l) and dihydrotestosterone (DTH, nmol/l) were measured in 54 children and young adults with male pseudohermaphroditism (46XY, no defect of steroid biosynthesis) 4 h after im injection of testosterone propionate (25 mg/m2, group 1, N = 18), or before and 2, 4 and 6 days after hCG (5000 IU/m2 im, group 2, N = 36). The response to hCG was also studied in 5 control children (unilateral cryptorchidism, group 3) and that to testosterone propionate in a gonadectomized child with confirmed 5α-reductase deficiency. Mean T (133.1 ± 14.0, sem) and DHT (17.1 ± 2.6) in group 1 were higher than in group 2 (17.3 ± 2.1 and 2.9 ± 0.4), but there was not significant difference in the T/DHT ratios (group 1: 10.7 ± 2.0; group 2: 7.2 ± 0.6). Following testosterone-propionate, there was a negative correlation of T with age (r = −0.723). After hCG, T and DHT were lower in the prepubertal children than in those under 2 or over 10 years, and the T/DHT-ratio rose with age. Two children from group 1 had a T/DHT-ratio above 18, but urinary aetiocholanolone/androsterone (Ae/A) ratios were normal. In the child with 5α-reductase deficiency, the T/DHT ratio was 60, and the urinary Ae/A ratio high. We concluded that the two tests are suitable for confirming or excluding 5α-reductase deficiency in prepubertal children, in whom basal DHT is too low for evaluation, but that physiological age-related changes in 5α-reductase activity have to be taken into consideration.