The association between corticosteroid treatment and gastric ulcer healing is controversial. The effects of corticosteroids on experimental ulcer healing in the rat were studied and the effect of coadministration of a prostaglandin E1 analogue was determined. Forty male adult rats were divided into five groups and pretreated for 10 days as follows: (a) control, (b) prednisolone (10 mg/kg), (c) prednisolone and misoprostol (300 micrograms/kg), (d) misoprostol, and (e) indomethacin (2 mg/kg) Gastric ulcer was induced by applying a cryoprobe to the serosal surface of the stomach. Healing was assessed by determining the ulcer size at three and six days. Mucosal regenerative activity at the ulcer edge was assessed by quantitating DNA synthesis in cells by immunohistochemical techniques using monoclonal antibodies to detect expression of proliferating cell nuclear antigen (PCNA) and incorporated 5-bromo-5-iododeoxyuridine (BrdU). Compared with control rats, the rate of healing and the mucosal regenerative activity were significantly reduced in rats treated with prednisolone and indomethacin (p < 0.05). Coadministration of misoprostol and corticosteroids reversed the delay in healing and the reduction in mucosal regeneration induced by corticosteroids alone. With misoprostol alone, the ulcer size and the number of epithelial cells that actively synthesised DNA did not differ from control animals. A comparison between the two immunohistochemical markers of cell proliferation showed a highly significant correlation between the two techniques (r = 0.64, p < 0.005), indicating that the simpler PCNA technique should prove valuable in assessing regeneration in experimental ulcer disease.