Insulin-like growth factor binding protein-3 (IGFBP-3) is an important modulator of the anabolic and mitogenic actions of the insulin-like growth factor (IGF) peptides. Previous studies have shown that the IGFs themselves can elevate levels of IGFBP-3 in vivo and in vitro. However, the regulatory mechanisms responsible for IGF-induced increases in IGFBP-3 are unclear. In this study we examined the expression of messenger RNA (mRNA) encoding IGFBP-3 in cultured bovine and human fibroblasts, two cell lines that secrete IGFBP-3 under IGF-I control. Northern analysis of bovine fibroblast RNA hybridized with a specific bovine IGFBP-3 complementary DNA probe indicated a single 2.8-kilobase (kb) transcript readily detectable within 2 h in IGF-I- or insulin-treated, but not in untreated, cells. IGFBP-3 mRNA abundance was maximal around 6 h, and remained elevated after 24 h of treatment. Secreted IGFBP-3 protein appeared more slowly. By Western ligand blotting, IGFBP-3 was not detected in medium from bovine fibroblasts incubated with IGF-I for 2, 4, or 6h, but was apparent after 24 h IGF-I treatment. Induction of IGFBP-3 mRNA was blocked when RNA synthesis was inhibited by actinomycin D. Furthermore, IGFBP-3 mRNA and protein was induced by different IGF-I analogs in direct relation to the ability of the peptides to bind to the type I IGF receptor, indicating a receptor-mediated process. GH had no effect on IGFBP-3 mRNA or protein levels in these cells. In contrast to its effect in bovine fibroblasts, IGF-I had no significant effect on steady state levels of IGFBP-3 mRNA in cultured human fibroblasts. A human IGFBP-3 complementary DNA probe hybridized to a single 2.8-kilobase mRNA species abundant in normal and SV40-transformed human fibroblasts under all culture conditions, and IGFBP-3 protein was secreted by these cells in the absence of exogenous stimuli. In human fibroblast cultures, IGF-I rapidly increased levels of IGFBP-3 in the medium without influencing transcript levels. Steady state levels of induced or constitutively expressed IGFBP-3 mRNA did not change significantly after 6h in the presence of actinomycin D, even though general RNA synthesis was inhibited more than 98%. These data demonstrate that expression of mRNA encoding IGFBP-3 is differentially controlled by IGF-I in bovine and human fibroblasts. Whereas cultured human fibroblasts may be suitable to study posttranscriptional regulation of IGFBP-3 availability, cultured bovine fibroblasts may provide a useful model system to probe the molecular mechanisms of IGFBP-3 gene expression and regulation by IGF-I.