Ninety-six children received morphine 0.1 mg/kg (n = 47) or ketorolac 1 mg/kg (n = 49) intravenously (IV) in a prospective, randomized, double-blind fashion, after tonsillectomy. Recovery variables and complications were recorded while subjects were in the hospital and parent(s) were contacted 24 h and 14 days after surgery. There were no differences in demographics, surgical management, awakening time, oxygen requirements, or time to readiness for postanesthesia care unit (PACU) discharge or discharge home between the two groups. Ketorolac subjects had fewer emetic episodes than morphine subjects (median 1 vs 3; P = 0.006) and were less likely to have more than two episodes of emesis after PACU discharge (9/49 vs 22/47; P = 0.007). Ketorolac subjects had more major bleeding (bleeding requiring intervention; 5/49 vs 0/47, one-tailed P = 0.03) and more bleeding episodes (0.22 episodes/subject vs 0.04 episodes/subject, P < 0.05) in the first 24 h after surgery, but no greater overall incidence of bleeding than the morphine subjects. In children having tonsillectomy, ketorolac, compared to morphine, reduced the number of emetic episodes after PACU discharge, but did not hasten awakening, readiness for PACU discharge or discharge home, and increased the likelihood of major bleeding in the first 24 h after surgery.