Gynaecomastia is frequent in pubertal boys and is regarded as a self-limiting abnormality. However, longitudinal studies proving this hypothesis are scarce.Longitudinal follow-up study (median 2.4, range 1.0-4.8 years).The regression of breast diameter was analyzed in 31 pubertal boys aged 11.7-16.1 (median 13.2) years with gynaecomastia. Furthermore, weight changes (as BMI-SDS) and pubertal stage, estradiol [E2], estriol, estrone, androstenedione, testosterone [T], dihydrotestosterone), gonadotropins, IGF-1, and IGFBP-3 serum concentrations determined at first clinical presentation were related to breast diameter regression determined by palpation and disappearance of breast glandular tissue in ultrasound in follow-up to identify possible predictors of breast regression.During the observation period the breast diameter decreased (in median -1 (interquartile range [IQR] -5 - +1) cm). At follow-up, 6% of boys had no breast enlargement any more, 65% developed lipomastia. Gynaecomastia was still present in 29%. None of the analyzed hormones was related significantly to breast diameter regression or disappearance of breast glandular tissue. In multiple linear regression analyses adjusted for observational period, as well as age and BMI-SDS at first presentation, changes in BMI-SDS (b-coefficient 6.0 ±2.3, p=0.015) but not the E2/T ratio or any other hormone determined at baseline was related to changes in breast diameter.Breast diameter regression seems not to be predictable by a hormone profile in pubertal boys with gynaecomastia. In pubertal boys presenting with gynaecomastia, conversion to lipomastia of smaller volume is common. The reduction of weight status was the best predictor of breast diameter regression.