Progesterone (P), 17-OH-progesterone (17-OH-P), androstenedione (.DELTA.4), dehydroepiandrosterone (DHEA) and testosterone (T) were determined in spermatic venous blood of 34 varicocele patients and of 13 normal subjects. The DHEA/.DELTA.4 ratio was also as an index of the .DELTA.5/.DELTA.4 pathway ratio in testosterone biosynthesis. The mean of T and .DELTA.4 in the spermatic blood of varicocele (V) patients appeared to be significantly lower with respect to that of normal (N) subjects (T:N = 1718.2 .+-. 202.4 (SEM [standard error of the mean]) nmol/l, No. 11; V = 1243.7 .+-. 97 (SEM) nmol/l, No. 34; P < 0.03. .DELTA.4: n = 56.4 .+-. 5.6 (SEM) nmol/l, No. 12; V = 38.1 .+-. 4 (SEM) nmol/l, No. 27, 0.02 > P > 0.01). A negative correlation was observed between the individual age of varicocele patients and 17-OH-P (No. 34, y = -30.66x + 1300, r = -0.57, P < 0.01) .DELTA.4 values (No. 27, y = -1.981x + 96.52, r = -0.67, P < 0.01). When the ratio of T precursors was evaluated, a positive correlation between the P/17-OH-P ratio and age of varicocele (No. 33, y = 0.0065x-0.092, r = 0.45, P < 0.03) was observed. The 17-OH-P/.DELTA.4 ratio was greatly increased with respect to that of normal subjects (n = 5.12 .+-. 0.93 (SEM), No. 12; V = 10.77 .+-. 1.31 (SEM), No. 27; P < 0.01). These data suggest that the reduced T levels in spermatic venous blood of varicocele patients were due firstly to the enzymatic deficiency of 17-20-lyase and secondly to that of 17.alpha.-hydroxylase activity as the patients grow relatively older. The negative correlation between the DHEA/.DELTA.4 ratio and .DELTA.4/T ratio was observed in normal subjects (No. 10, y = -0.00432x + 0.0542, r = -0.67, P < 0.03) as well as in varicocele patients (No. 27, y = -0.00399x + 0.0587, r = -0.48, 0.02 > P > 0.01). This indicates that in the testis of varicocele patients the testosterone remains prevalently supplied by the .DELTA.5 pathway of biosynthesis.