Although differences in plasma GnRH concentrations have been identified after iv and sc injection of this peptide, differences in pituitary LH responses to iv and sc GnRH have not been evaluated in detail. We studied the magnitude and contour of plasma GnRH and LH responses after low doses of iv and sc GnRH administered to men with idiopathic hypogonadotropic hypogonadism and compared them to LH pulses in normal men after endogenous GnRH secretion. Mean areas under the LH response curves differed significantly (P less than 0.01) after 25 ng/kg, but not 250 ng/kg, iv and sc GnRH doses. The mean time from basal to peak plasma LH concentrations was significantly longer with sc than iv GnRH (P less than 0.02). In addition, individual LH responses were more variable with sc GnRH. Intravenous administration produced greater GnRH amplitude (P less than 0.001) and area under the curve (P less than 0.005) and shorter time to peak (P less than 0.01) GnRH concentrations. When plasma LH responses of similar area and amplitude were compared, the contour of LH responses after iv GnRH more closely simulated the LH pulses in normal men. These data demonstrate that 1) significant differences exist in the amplitude, contour, and variability of plasma LH and GnRH pulses after iv and sc GnRH; and 2) iv GnRH elicits LH secretory episodes which closely resemble endogenous pulsations of normal men. These results suggest that iv GnRH administration may be preferred in physiological studies and, if the data can be extrapolated to women, may account for the greater success of ovulation induction reported with iv GnRH.