The literature presented supports a small but highly probable association between corticosteroids and ulcers. The following characteristics appear to be exhibited by patients who are at high risk for developing corticosteroid-induced ulcers: corticosteroids coadministered with NSAIDs, a total dosage greater than 1000 mg of prednisone equivalent, a duration of therapy longer than 30 days, and a history of PUD. Further prospective research examining the association of corticosteroids and PUD in conjunction with other contributing factors is needed. The role of prophylactic therapy to prevent corticosteroid-induced ulcers is not well established. Even though a small study in liver transplant recipients, who are already at increased risk for GI ulceration, has suggested beneficial effects with prophylactic regimens, generalization of these results to all corticosteroid-treated patients would be inappropriate. Large prospective trials to determine the most efficacious prophylactic regimen (e.g., histamine2-receptor antagonists, proton pump inhibitors, cytoprotective agents [misoprostol]) for corticosteroid-induced ulcerations are not currently available. We suggest that most prophylaxis currently performed is unnecessary. In high-risk patients, however, prophylaxis appears to be prudent until further information is available.