OBJECTIVE Gastrointestinal symptoms, particularly pyrosis, complicate nonsteroidal anti-inflammatory drug (NSAID) use. NSAIDs cause esophageal injury, and H2 blockers are often prescribed for, and successfully control, NSAID-related symptoms. To determine whether NSAIDs can induce gastroesophageal reflux, we studied the effect of a commonly used NSAID, naproxen, on reflux parameters and esophageal function. METHODS Nine healthy volunteers (five males, four females, age 23-34 yr) were studied. After basal measurements were taken, the subjects randomly received naproxen 500 mg p.o. b.i.d. or placebo for 1 wk. On day 6, the subjects underwent esophageal manometry with a water-perfused system and Dent sleeve. Body pressures, contraction velocity, and duration of contraction were recorded in the distal 7 cm of the esophagus. The lower esophageal sphincter pressure (LESP) and number of transient relaxations (TLESRs) were monitored. This was followed by 24-h pH monitoring. The subjects then crossed over to the other drug after a minimum 14-day wash-out period. RESULTS No subject experienced any GI symptoms during the study. One subject developed reflux-induced symptoms a few months after completing the study and was excluded from the analysis. The total fraction of time (pH < 4) was 4.9 +/- 1.0% in the basal state, 5.5 +/- 1.4% on placebo, and 5.4 +/- 1.5% on naproxen. These differences were not significant. The number of reflux episodes and the esophageal clearance time were not affected by naproxen. The LESP in the basal state was 32.1 +/- 5.6 mm Hg, 32.3 +/- 4.2 mm Hg on placebo, and 29.9 +/- 3.3 mm Hg on naproxen (p = NS). The number of TLESRs per 30 minutes in the basal state was 3.5 +/- 0.9, 4.6 +/- 1.2 on placebo, and 5.8 +/- 1.0 on naproxen (p = NS). The speed and duration of contractions were not affected by naproxen. The excluded subject had marked basal reflux (total fraction of time pH < 4 = 10.7%), low LESP (8 mm Hg), and a marked increase in reflux on naproxen (total fraction of time pH < 4 = 53%). CONCLUSIONS Naproxen did not induce reflux in normal subjects, although reflux did increase in some subjects. Naproxen had no significant effect on motility parameters. Our data suggest that NSAIDs do not impair the anti-reflux barrier or induce reflux. Pyrosis experienced during NSAID use may not arise from the esophagus or may reflect altered esophageal sensitivity. A single subject with decreased LESP and asymptomatic increased acid exposure in the basal state had a marked increase in reflux on naproxen. This person subsequently developed symptomatic gastroesophageal reflux. The effect of NSAIDs on individuals with a propensity to reflux deserves further study.