The authors compared diurnal growth hormone (GH) secretion with GH secretion during sleep in 24 children with delayed growth. In group I (children with normal response to provocative tests), the level of daytime secretion was lower than that of nocturnal secretion. In 3 of 9 cases, daytime secretion was abnormal, whereas nocturnal secretion was normal. In 2 cases, both diurnal and nocturnal secretion were abnormal, but response to provocative stimuli was normal. In group II (children with a false partial GH deficiency, i.e. with inadequate response to provocative tests, GH peak < 11 ng/ml and normal nocturnal secretion), the results were comparable with those of group I, with extremely low diurnal secretion in 6 of 9 cases. In group III (children presenting true partial GH deficiency, i.e. GH < 11 ng/ml in response to provocative tests together with abnormal nocturnal secretion), both diurnal and nocturnal GH secretion were insufficient, with nonexistent diurnal secretion in 5 of 6 cases. Diurnal secretion does not seem to be a reliable indicator of 24-hour spontaneous secretion.