The aim of this study was to determine the effectiveness of highly selective vagotomy (HSV) or misoprostol, a prostaglandin E1 (PGE1) analog, for protecting the gastroduodenal mucosa (GDM) from the effects of diclofenac sodium (DS). Fifty mongrel dogs were randomly allocated to five groups. HSV alone was performed in group I dogs (controls) to standardize the operation. DS was given intramuscularly for 12 consecutive days to the group II dogs, whereas in the group III dogs HSV was performed, followed a month later by DS administration, as in group II. DS was given in combination with misoprostol for 12 days to the group IV dogs. HSV was performed on the group V dogs, and a month later DS and misoprostol were given, as in group IV. After sacrificing the animals the GDM was examined for macroscopic and histologic lesions. Statistical analysis was made by Fisher's exact test. HSV alone did not protect the gastric or duodenal mucosa from the effects of DS (p = 0.474 and p = 0.62, respectively). Misoprostol alone also did not offer significant protection to the gastric or the duodenal mucosa (p = 0.08 and p = 0.65, respectively). The combination of HSV plus misoprostol protected the gastric mucosa (group V, p = 0.007) but not the duodenal mucosa (group V, p = 0.08). Hence HSV or misoprostol alone offers no protection to the GDM from the effects of DS. The combination of HSV and misoprostol offers significant protection only to gastric mucosa. Enhancement of the mucosal defense mechanisms combined with strong reduction of gastric acidity may offer adequate protection to gastric mucosa from the effects of nonsteroidal antiinflammatory drugs.