OBJECTIVE Growth hormone (GH) secretion is increased in conditions of GH insensitivity such as Laron syndrome, with elevation of both basal and peak levels. We have studied the effect of recombinant IGF-I therapy on the pattern of GH secretion in two subjects with GH insensitivity. METHODS Two pubertal subjects with GH insensitivity (female, 16.4 years, breast stage 3; male 13.6 years, genital stage 2) were investigated after 6 months of IGF-I therapy (120 micrograms/kg twice daily s.c. at 0800 and 1900 h). GH profiles taken before the start of IGF-I therapy, when both subjects were prepubertal (aged 14.0 and 11.5 years respectively), were used for comparison. METHODS GH profiles were performed with blood samples taken every 20 minutes between 2000 and 0800 h from an indwelling cannula. METHODS Serum samples were assayed for GH by immunoradiometric assay and IGF-I, IGFBP-1 and insulin by radioimmunoassay. RESULTS Before IGF-I therapy, GH profile studies demonstrated pulsatile GH secretion. Basal GH was elevated with no value falling below the limit of detection of the assay and an increase in peak levels (maximum 203 and 206 mU/l at 0000 h and 0020 h respectively). After 6 months IGF-I therapy, the GH profiles were significantly different. With the onset of puberty a further increase in GH secretion would have been expected; nevertheless, following administration of IGF-I at 1900 h, GH secretion decreased with a reduction in mean overnight GH levels from 65 to 33 mU/l and 53 to 11 mU/l respectively. GH pulsatility was also suppressed in the two subjects, for the first 3.5 and 6 hours overnight respectively. Pulsatile GH secretion then returned with peak levels reaching 130 and 63 mU/l respectively. Prior to therapy IGF-I levels were at the lower limit of assay detection. On IGF-I therapy serum IGF-I levels reached a peak within 3 hours (298 and 438 micrograms/l) coinciding with the suppression of GH secretion. IGF-I levels fell rapidly overnight to 92 and 101 micrograms/l at 0800 h prior to the next injection. The fall in serum IGF-I coincided with the return of GH secretion. IGFBP-1 levels increased overnight both before and during IGF-I therapy, rising from 24 to 83 and 22 to 110 micrograms/l before therapy and 13 to 60 and 13 to 71 micrograms/l during therapy. This rise in IGFBP-1 appeared to be inversely related to the fall in serum insulin levels overnight and appeared not to be affected by IGF-I therapy. CONCLUSIONS GH secretion is suppressed by exogenous IGF-I therapy in GH insensitive subjects. The failure to maintain high serum IGF-I levels overnight, presumably due to a persisting defect in serum IGFBP-3 levels, was associated with an early return of GH secretion. These findings may have implications for the dose and regimen of IGF-I therapy in subjects with growth hormone insensitivity.