A review of the literature is presented on the gastrointestinal effects of etodolac, a new nonsteroidal anti-inflammatory drug (NSAID), as evaluated in both microbleeding and endoscopic studies. In four microbleeding studies, gastrointestinal blood loss in healthy subjects was estimated by a 51Cr-erythrocyte labeling method before drug treatment, after 7 days of treatment with NSAIDs including etodolac, and 1 week after the last day of treatment. In these 7-day studies, the gastrointestinal blood loss seen with etodolac (600 to 1200 mg/day) was similar to that seen with placebo and significantly (p less than 0.05) less than that seen with aspirin (2600 mg/day), naproxen (750 mg/day), ibuprofen (2400 mg/day), or indomethacin (200 mg/day). Naproxen, ibuprofen, and indomethacin caused mean daily blood losses in excess of 1 ml/day over baseline values. The increase with aspirin was 4 to 5 ml/day. In contrast, the greatest mean daily increase in blood loss with etodolac therapy was 0.2 ml. In a 4-week study of etodolac (600 and 1000 mg/day) and piroxicam (20 mg/day) given to patients with osteoarthritis or rheumatoid arthritis, blood loss seen with etodolac was comparable to that seen with placebo and significantly less than that seen with piroxicam. Gastrointestinal irritation was also assessed by endoscopy after 1 week of NSAID or placebo treatment. Endoscopy scores after etodolac treatment (up to 1200 mg/day) were similar to scores at baseline and after placebo and were significantly lower than scores following treatment with aspirin (3900 mg/day), indomethacin (200 mg/day), ibuprofen (2400 mg/day), or naproxen (100 mg/day). The effects of etodolac (600 or 1000 mg/day) and diclofenac (150 mg/day) were not different from each other or from baseline. These data indicate that etodolac, in these studies, did not cause clinically significant gastrointestinal microbleeding or visible gastric injury. By the criteria used in these studies, etodolac is less irritating to the gastrointestinal tract than aspirin, indomethacin, ibuprofen, naproxen, or piroxicam, and compares favorably with diclofenac.