The aim of this study was to evaluate the usefulness of basal measurements of gonadotropins in distinguishing between constitutionally delayed puberty (DP) and hypogonadotropic hypogonadism (HH), comparing its diagnostic efficiency with that of the dynamic GnRH infusion test (0.83 microg/min during 120 min). We studied 20 males, chronological age (CA) 14-18 years, with a final diagnosis of DP (n = 8), partial HH (n = 5) and complete HH (n = 7), confirmed by follow-up. We also evaluated basal samples of ultrasensitive LH and FSH in 117 healthy control males (CA 2-19 yr), classified according to Tanner stage. In the control group, ROC plot analysis showed a cutoff to differentiate prepuberty from puberty of 0.65 IU/l for LH (sensitivity: 91%, specificity: 98%). Differences were found (p < 0.05) in basal LH and in maximal responses to GnRH in complete HH in relation to DP and partial HH. The diagnostic efficiency of the GnRH infusion test was 85%. For basal LH, a cut-off limit of 0.65 IU/l showed a diagnostic efficiency of 85% for complete HH and 100% for partial HH and DP. We conclude that, in our experience, basal LH levels above 0.65 IU/l measured by ultrasensitive assay would rule out a complete deficiency. It was not possible to differentiate DP from partial HH, either in basal samples or with the infusion test.