Nonsteroidal antiinflammatory drugs (NSAIDs) have been linked with a high incidence of ulcer complications. Histologic gastritis is present in most patients with standard peptic ulcers, and this gastritis is generally related to Helicobacter pylori (HP). We questioned whether gastric ulcers associated with nonsteroidal antiinflammatory drugs (NSAIDs) develop via a novel mechanism, distinct from the usual HP-gastritis-ulcer diathesis. Two groups of patients with newly discovered gastric ulcers were assessed: 1) daily NSAID use > 1 month (n = 19), 2) no NSAID use (n = 36). Biopsy specimens from the rim of the ulcer and adjacent normal mucosa were coded, randomized, and evaluated for histologic features and HP. HP prevalence was significantly lower in the NSAID group (10/19 (53%) vs. 30/36 (83%), p = 0.01). In biopsies from the ulcer rim, inflammatory cell density and epithelial abnormalities were significantly less in the NSAID group than in the no-NSAID group. Biopsies from adjacent mucosa exhibited the same trends, but the differences did not reach statistical significance (inflammation, p = 0.06; epithelium, p = 0.05). HP-positive patients had similar inflammation and epithelium scores, whether or not they took NSAIDs. However, HP-negative NSAID patients had significantly lower scores than HP-positive NSAID users (inflammation, 0.6 +/- 0.2 vs. 2.4 +/- 0.4; epithelium, 1.1 +/- 0.6 vs. 3.5 +/- 0.7). Patients with NSAID-associated gastric ulcers have a lower prevalence of HP and less histologic gastritis than patients with non-NSAID gastric ulcers. The gastritis is related to the underlying HP and not to NSAID ingestion. NSAID-associated gastric ulcers may represent a major subset of peptic ulcers that do not require HP for their development.