We recently identified and characterized a protease present in rat serum which is capable of generating des(1–3)IGF-I. In this study, we have investigated the effects of GH deficiency and replacement on the activity of this protease in rat serum and tissue extracts. Protease activity was significantly higher in sera from hypophysectomized (hypox) rats than sham-operated rats (P<0·001) and GH treatment of hypox rats (human GH, 100 μg/100 g body weight i.p. for 10 days) significantly reduced the levels towards normal. The addition of IGF-I to hypox rat serum to achieve IGF-I concentrations comparable with or greater than that seen in normal rat serum had no effect on the measured protease activity. Protease activity was also detected in tissue extracts. The level of protease activity in the various tissues from sham-operated rats demonstrated the following order: liver>testes>heart> skeletal muscle>lung>thymus>kidney>brain>spleen. In all tissue extracts examined, except that from the lung, the levels of protease activity were higher in extracts from hypox rats compared with sham-operated rats. The largest differences between tissue extracts from hypox and shamoperated rats were seen in spleen (4-fold higher), kidney (2·27-fold), testes (1·55-fold) and heart (1·31-fold). In the liver, kidney and testes, GH treatment significantly reduced protease activity. Since the pattern of serum IGF-binding proteins (IGFBPs) differ in hypox rats compared with normal rats, we determined whether these changes could result in enhanced serum binding of des(1–3)IGF-I. Sera from hypox, GH-treated hypox and sham-operated rats were incubated with either 125I-IGF-I or 125I-des(1–3)IGF-I and analyzed by Sephacryl S-200 chromatography. Under the conditions used, 125I-des(1–3)IGF-I was recovered exclusively in the 7·5 kDa peak when incubated with serum from sham-operated rats. After incubation with sera from hypox or GH-treated hypox rats approximately 20% of the 125I-des(1–3)IGF-I was recovered in the 50 kDa peak with the remaining radioactivity recovered as free peptide. These observations suggest that the generation of des(1–3)IGF-I may be inversely regulated by GH and may represent yet another site of GH regulation of IGF-I action. The changes in the pattern of serum IGFBPs which occur in GH deficiency are unlikely to negate the enhanced generation of des(1–3)IGF-I by increasing binding capacity for this IGF-I variant.