The aim of this study was to explore pulp healing and reparative dentinogenesis following pulp-capping by using recombinant human insulin-like growth factor I (rhIGF-I). Exposures were made through the mesial pulp horn in first upper molars in two-month-old Wistar rats. The pulp was covered with one dose of sterile 4% methylcellulose gel containing either 400 ng rhIGF-I or saline in contralateral controls. The exposure site was closed with sterile Teflon membrane, and the cavity was filled with IRM cement. Additional molars were capped with Dycal as controls. After 3, 7, or 28 days, animals were anesthetized and fixed by intravascular glutaraldehyde perfusion. Molars were decalcified and processsed for histological analysis and cut with membrane and residual methacrylate from IRM in situ. Only specimens with acceptable pulp sealing according to blinded microscopy control were included. On day 3, identical inflammatory responses in the upper pulp were observed in molars with rhIGF-I gel or control gel. On day 7, granulation tissue ingrowth had partly replaced inflammatory infiltration in both groups. After 28 days, complete dentin bridging and tubular dentin formation were observed more frequently and closer to the test substance containing rhIGF-I. The reparative dentin response to capping with rhIGF-I was similar to that after the use of Dycal. In conclusion, microscopic control of membrane sealing in situ gives valid information on the more subtle pulp effects of growth factors. The observations suggest that pulp-capping of rat molars by means of rhIGF-I enhances reparative dentinogenesis in comparison with vehicle controls.