Flunixin meglumine has been reported to induce gastrointestinal lesions in dogs when administered at therapeutic dosages. We administered flunixin meglumine to dogs daily for 10 days to assess the effect of this drug on the gastrointestinal tract. We also evaluated the possibility of corticosteroid potentiation of gastrointestinal toxicosis by concurrent administration of prednisone to 1 group of dogs. Dogs were monitored for gastrointestinal toxicosis by means of serial endoscopic evaluation, measurement of fecal occult blood, PCV, and total solid concentration, and by physical examination. There were 3 treatment groups of 5 dogs each. Group-1 dogs were given 2.2 mg of flunixin meglumine/kg daily, in 2 divided doses IM; group-2 dogs were given 4.4 mg of flunixin meglumine/kg daily, in 2 divided doses IM; and group-3 dogs were given 2.2 mg of flunixin meglumine/kg daily, in 2 divided doses IM plus 1.1 mg of prednisone/kg/d orally, in 2 divided doses. A fourth group of 5 dogs served as a control group. Endoscopically visible gastric mucosal lesions developed in all treated dogs within 4 days of initiating treatment. Lesions first developed in the gastric pylorus and antrum and lesions at these sites were more severe than those observed elsewhere. Dogs treated with flunixin meglumine plus prednisone developed the earliest and most severe lesions; lesion scores in group-2 dogs were higher than those in group-1 dogs. All dogs treated had occult blood in their feces by day 5 and its presence appeared to correlate more closely with endoscopic findings than did physical examination findings or changes in values for PCV or total solids.(ABSTRACT TRUNCATED AT 250 WORDS)