The basal and gonadotropin-releasing hormone (GnRH) stimulated levels of LH were measured in 21 boys with delayed puberty using conventional RIA, mouse interstitial cell in vitro bioassay (B-LH), and a highly sensitive immunofluorometric method (F-LH). On the basis of subsequent clinical follow-up, the subjects were diagnosed as idiopathic constitutional delay of puberty (CD, n = 13) or hypogonadotropic hypogonadism (HH, n = 8). The basal RIA LH levels were similar in the two diagnostic groups (HH, 2.92 ± 0.76, CD 3.53 ± 1.37 IU/L). In contrast, the mean basal B-LH was significantly lower in boys with HH than with CD (1.10 ± 0.45 versus 2.91 ± 1.23 IU/L; p < 0.01). A similar finding was made by F-LH measurements which were clearly lower in the HH than the CD group (0.073 ± 0.04 versus 1.71 ± 0.97 IU/L, p < 0.01). Also upon GnRH stimulation (3.5 Mg/kg i.v.), the distinction between the CD and HH groups was better with the B-LH and F-LH measurements. The basal B/I ratio of the CD group (0.90 ± 0.43) was more than that of the HH group (0.42 ± 0.25, p < 0.01) and this ratio increased significantly (more than 2-fold, p < 0.01) during GnRH stimulation in the CD group, but not in the HH patients. Such differences were not found between the B/F ratios of the CD and HH groups. Measurements of basal and GnRH stimulated B- and F-LH levels clearly improved the distinction between CD and HH in comparison to the conventional RIA method, due to the low sensitivity and likely cross-reactions with some non-LH constituents of serum in the latter assay. This problem is, to a great extent, eliminated by better sensitivity and specificity of B-LH and F-LH measurements. For the same reasons, the difference in B/I ratios between the CD and HH samples, and the increased B/I ratio after GnRH stimulation in the CD group, were not observed in B/F ratios. In conclusion, the measurements of basal and GnRH-stimulated concentrations of serum B-LH and F-LH clearly improve the differential diagnostics between CD and HH. The discrepancies measured between the B/I and B/F ratios in these samples call for reevaluation of the bio/immuno ratios of circulating LH. © 1990 International Pediatric Research Foundation, Inc.