Six normal stallions of light horse breeds aged 5-17 years were used from fall to winter to investigate the difference between steroid hormone concentrations in testicular and jugular venous blood before and after exogenous GnRH. At 48 h before experimentation, an indwelling cannula was placed surgically in the testicular vein of the stallion. After the stallion recovered from anaesthesia, a catheter was placed percutaneously in the jugular vein. Each stallion was housed in a tie stall to allow simultaneous sampling of jugular or testicular blood. On the first and second sampling days, respectively, 1 ml of physiological saline solution and a 1 ml solution of GnRH (25 micrograms) were administered intravenously. Samples were taken from both sites at intervals from 60 min before treatment to 780 min after treatment. Plasma was analyzed for luteinising hormone (LH) and follicle-stimulating hormone (FSH), 17 beta-hydroxyandrogens (androgens), oestrone and oestrogen conjugates by radioimmunoassay. Pre-treatment (baseline) plasma concentrations of both LH and FSH between jugular and testicular samples were similar. The difference between basal levels of jugular and testicular androgens, oestrone and oestrogen conjugates were 144-fold, 60-fold and 13-fold respectively, although individual variation was observed. A low dose of exogenous GnRH produced a significant LH and FSH response in testicular and jugular plasma (P less than 0.05). There were no significant changes in steroid secretion caused by the increases in LH and FSH (P greater than 0.05), although individual variation in the androgen response was apparent (P less than 0.1). There was a positive correlation between basal testicular venous androgen levels and the magnitude of the FSH response to GnRH (P less than 0.05). Significant correlations between baseline oestrogens and the magnitude of the gonadotrophin response was not observed. Surgery depressed jugular oestrogen conjugate values (P less than 0.001) when compared to pre-surgical samples. Spermatogenesis also was depressed (P less than 0.01) by surgical manipulation, although total viable spermatozoa counts returned to normal limits within 3-5 months post operatively. We developed a model that allows the study of dynamic endocrine events associated with the hypophyseal-gonadal axis of the stallion. Our findings confirm the presence of a testicular-jugular hormone gradient in the unanaesthetized stallion. We have demonstrated that a relatively low dose of GnRH can induce a significant gonadotrophin response and a variable androgen response, but not a significant oestrogen response. Although baseline levels of androgens and not oestrone and oestrogen conjugates appeared to affect pituitary responsiveness, other steroidogenic components may be involved.(ABSTRACT TRUNCATED AT 400 WORDS)