5-(4-Chlorobenzoyl)-1,4-dimethyl-1H-pyrrole-2-acetate dihydrate (Zomepirac, Zomax) was assessed in a variety of preparations in which narcotic agonists and antagonists have distinctive profiles. Zomepirac failed to displace tritiated etorphine significantly at concentrations up to 200 mumol/l. It was active (3-7 X 10(-5) mol/l) on both the electrically stimulated guinea pig ileum and mouse vas deferens but was less efficacious than morphine; these effects were not reversed by narcotic antagonists. Zomepirac was inactive in some analgesic assays in mice: tail-flick (1-30 mg/kg s.c.); hot plate (1-100 mg/kg s.c.); Nilsen (1-50 mg/kg s.c.); it was also inactive (1-10 mg/kg s.c.) as an antagonist of morphine in the tail-flick assay. It was, however, active in the paraphenylquinone writhing assay; this action was not reversed by naloxone (1-10 mg/kg) Zomepirac (2.5-10 mg/kg s.c.) failed to suppress signs of withdrawal produced by morphine deprivation in dependent rhesus monkeys. The intraperitoneal infusion of high doses of zomepirac (50 mg/kg/24 h or more) in rats produced toxicity. However, lower doses failed to induce narcotic-like dependence over a 6-7-day infusion period. The compound failed to maintain intravenous drug self-administration responding in rhesus monkeys over a range of doses (0.1-3.2 mg/kg) although codeine did so. In rhesus monkeys trained to discriminate narcotic agonists (etorphine or ethylketazocine) from saline, zomepirac (1-10 mg/kg i.m.) failed to produce drug-appropriate responding. Thus, zomepirac appears to possess few, if any, of the characteristic actions of narcotics that are associated with their abuse liability.