Anti-ulcer prostaglandins (PG)--misoprostol, enprostil and rioprostil--have been given to more than 5000 patients in short-term studies on gastric and duodenal ulcer. Analysis of these studies shows the drugs to be safe. Their side effects appear to be dose-dependent and mainly restricted to the gastrointestinal system, the major syndromes being diarrhoea and abdominal pain. The clinical relevance of PG-related unwanted effects, though in average exceeding that of H2-blockers, seems to be sufficiently low. In terms of safety efficacy, however, they appear inferior to H2-antagonists, so their routine use in preference to the latter compounds is still premature.