NSAID use and Helicobacter pylori both cause damage to the gastric mucosa and can cause peptic ulcers. Our aim was to test the relationship between gastric mucosal polymorphonuclear leukocyte (PMN) infiltration and the severity of NSAID-induced gastric injury. H. pylori density, mucosal interleukin-8 (IL-8), and nitrite levels were assessed after receiving placebo and again after receiving 1000 mg of naproxen daily for three days. Histology was graded using a visual analog scale (0-5). IL-8 levels were assayed by ELISA and nitrite levels by Griess reaction. Eleven healthy volunteers with H. pylori infection entered. All had normal-appearing gastric mucosa after placebo. Postnaproxen gastric damage included three with none, one with mild, three with moderate, two with severe, and three were very severe mucosal injury (including one with an ulcer >5 mm). There was an inverse correlation between endoscopic score and the pH of the gastric juice post-therapy (R = -0.77, P = 0.004). There was no significant change in histologic or biochemical parameters from pretreatment levels. And none of the parameters (eg, PMN density) predicted endoscopic outcome. In conclusion, there was no relation between mucosal PMN density and endoscopic mucosa injury. PMN infiltration, while not predictive, may be a surrogate for an H. pylori infection-related increased risk of NSAID ulcers.