Two hundred forty patients with benign gastric ulcer were treated in a controlled clinical trial to assess the effect on healing of cimetidine, antacids, and hospitalization. Inpatients and and outpatients were randomly assigned to one of three treatments: cimetidine plus antacid, cimetidine plus dummy antacid, or placebo tablet plus antacid. In 206 patients who met criteria for analysis, ulcer healing as shown by endoscopy occurred by 12 days in 11 to 26 percent and by 42 days in 58 to 76 percent. There were no significant differences in healing between hospitalized and nonhospitalized patients or between treatment subgroups. Symptomatic response was equivalent in all groups. The median antacid consumption was 328 mEq of in vitro buffering capacity per day. Patients taking antacids experienced significant diarrhea compared with those taking no antacid. This investigation suggests that the effect of cimetidine is equivalent to that of large amounts of antacid, but because a true placebo group was not studied it is not possible to conclude from this study alone whether either agent influenced healing. In contrast to widespread belief, initiation of treatment in the hospital did not enhance healing, but because patients were not randomly assigned to inpatient and outpatient status no final conclusion about the effect of hospitalization on healing can be drawn.