Gonadotropin-releasing hormone analogues (GnRH-A) induce inhibition of testicular function and reduction of serum testosterone (T) in man, but the mechanism involved is still debatable. To elucidate it we studied six patients with hypogonadotropic hypogonadism (HH) in chronic substitution with hCG for correction of androgen deficiency symptoms, and evaluated the effect of addition of GnRH-A to the hCG therapy on plasma levels of T and 17 alpha-hydroxyprogesterone (17 OHP). All patients were treated with 1000 U of hCG in every 3rd day for 24 weeks. After 8 weeks of this regimen, GnRH-A, Buserelin (D-Ser-TBU-EA-LHRH), 200 micrograms per day sc, was added and given for 8 weeks. After cessation of analogue administration patients were followed for 8 further weeks. The levels of the two steroids did not differ markedly in the pre- and post-GnRH-A period. GnRH-A given for two months did not lower T or 17 OHP levels as in eugonadal men after similar treatment. The median T concentrations during GnRH-A tended to be increased, with plasma values higher (P less than 0.025) than the peak values observed during hCG alone. Since administration of Buserelin did not inhibit hCG-sustained steroid levels in these HH patients, it is conceivable that GnRH-A may have lacked a direct inhibitory gonadal effect in such experimental conditions.