BACKGROUND: The use of dipyrone as an analgesic is controversial. It is used most commonly to treat postoperative pain, colic pain, cancer pain and migraine, and in many countries, eg, Russia, Spain, Brazil, and in many parts of South-America and Africa, it is the most popular non opioid first line analgesic. In others it has been banned (e.g. USA, UK) because of its association with potentially life-threatening blood dyscrasias such as agranulocytosis. Dipyrone is currently available in Austria, Belgium, France, Germany, Italy, The Netherlands, Spain, Switzerland, South Africa, Latin America, Russia, Israel and India. OBJECTIVES: To assess quantitatively the analgesic efficacy and adverse effects of single-dose dipyrone in randomised trials in moderate to severe postoperative pain. To compare the relative efficacy of dipyrone with other drugs assessed in the same way. SEARCH STRATEGY: Published reports were identified from Medline, Embase, the Cochrane Library (Issue 3 1999), LILACs and the Oxford Pain Relief Database. Additional studies were identified from bibliographies of retrieved reports. Date of the most recent search: December 1999. SELECTION CRITERIA: The following inclusion criteria were used: full journal publication, clinical trial, random allocation of patients to treatment groups, double-blind design, adult patients, pain of moderate to severe intensity at the baseline assessment, postoperative administration of study drugs, treatment arms which included dipyrone and placebo or active control and oral, rectal, intramuscular or intravenous administration of study drugs. DATA COLLECTION AND ANALYSIS: Summed pain intensity and pain relief data over 4-6 hours were extracted and converted into dichotomous information to yield the number of patients who obtained at least 50% pain relief. This was used to calculate the proportion of patients with, and number-needed-to-treat for, at least 50% pain relief over 4-6 hours. Single-dose adverse effect data were collected. MAIN RESULTS: Fifteen studies were included; eight used placebo and seven used an active control (oral dexketoprofen 12.5 mg or 25 mg, oral ketorolac 10 mg, intramuscular pethidine 100 mg or ketorolac 30 mg, intravenous tramadol 100 mg or rectal suprofen 300 mg). In five trials (288 patients) the mean response rate (proportion of patients with at least 50% pain relief) for single dose oral dipyrone 500 mg was 73% (range 54% to 87%) and with placebo it was 32% (19% to 41%) in moderate to severe postoperative pain over 4-6 hours. In two studies (113 patients) the response rate with oral dipyrone 1 g was 69% (61% and 77%) and with placebo it was 20% (11% and 25%). In one study (70 patients) the response rate with intramuscular dipyrone 2 g was 74% and with placebo it was 46%. No analyses could be conducted for adverse effects. The response rates in the active controlled trials were similar to those reported in the placebo controlled trials. REVIEWER'S CONCLUSIONS: Single-dose dipyrone appears to be of similar efficacy to ibuprofen 400 mg and other analgesics frequently used in the treatment of moderate to severe postoperative pain. The commonest adverse effects were somnolence, gastric discomfort and nausea.