Development of IFMA assays with 50-to 100-fold increased sensitivity compared to RIA allowed a more accurate assessment of changes in gonadotropin levels in normal puberty. We assessed the usefulness of an IFMA assay for LH and FSH in the differential diagnosis between central precocious puberty (CPP) and premature thelarche (PT), by studing 11 girls (12 tests) with PT in progression (CA, X±SD: 4.1±2.3y, BA: 4.3± 2.9y) and 5 girls with involutive thelarche (IT) (CA: 4.9±2.7y, BA: 6.0± 3.1y) at the moment of the test and 8 girls with CPP (CA: 6.8± 1.4y, BA: 9.3±1.9y). Diagnosis was done by clinical and auxological criteria, pelvic ultrasound and clinical follow up for at least one year after the test. Results of LH and FSH serum levels were specifically excluded as diagnostic criteria in this study. LH and FSH serum levels were measured by both RIA and IFMA methods before and 20, 30, 60′ after IV administration of a 25-μg bolus of GnRH. A ratio of maximal response of LH and FSH (Mx LH/FSH) was considered as pubertal reactivation of GnRH pulse generator when values were >1 for RIA and >0.3 for IFMA. (Arch Pediatr Adolesc Med 1994;148:369). Results: Basal and stimulated LH and FSH levels measured by RIA or by IFMA could not differentiate all PPC girls from girls with PT. Using RIA, Mx LH/FSH ratio>1 was observed in 6/8 GnRH tests in girls with CPP, however similar ratio values were also found in 2/5 girls with IT and in 3/12 girls with progressive PT. (Sensitivity: 75%, Specificity: 71%). All girls with IT as well as those with progressive thelarche showed Mx LH/FSH ratio ≤0.3 when samples were assessed by IFMA assays and Mx LH/FSH ratio >0.3 was observed in 7/8 girls with CPP. (Sensitivity: 88%, Specificity: 100%). Conclusion, the use of an ultrasensitive IFMA assay for measuring gonadotropin serum concentration in response to conventional GnRH challenge allows to improve the detection of an actual pubertal reactivation of the GnRH pulse generator.