This review examines the evidence from published data concerning the tolerability (indicated by the incidence of nausea, vomiting, sedation, pruritus, and urinary retention), of three analgesic techniques after major surgery; intramuscular analgesia (i.m.), patient-controlled analgesia (PCA), and epidural analgesia. A MEDLINE search of publications concerned with the management of postoperative pain and these indicators identified over 800 original papers and reviews. Of these, data were extracted from 183 studies relating to postoperative nausea and vomiting, 89 relating to sedation, 166 relating to pruritus, and 94 relating to urinary retention, giving pooled data which represent a total of more than 100,000 patients. The overall mean (95% CI) incidence of nausea was 25.2 (19.3-32.1)% and of emesis was 20.2 (17.5-23.2)% for all three analgesic techniques. PCA was associated with the highest incidence of nausea but the emesis was unaffected by analgesic technique. There was considerable variability in the criteria used for defining sedation. The overall mean for mild sedation was 23.9 (23-24.8)% and for excessive sedation was 2.6 (2.3-2.8)% for all three analgesic techniques (significantly lower with epidural analgesia). The overall mean incidence of pruritus was 14.7 (11.9-18.1)% for all three analgesic techniques (lowest with i.m. analgesia). Urinary retention occurred in 23.0 (17.3-29.9)% of patients (highest with epidural analgesia). The incidence of nausea and excessive sedation decreased over the period 1980-99, but the incidence of vomiting, pruritus, and urinary retention did not. From these published data it is possible to set standards of care after major surgery for nausea 25%, vomiting 20%, minor sedation 24%, excessive sedation 2.6%, pruritus 14.7%, and urinary retention requiring catheterization 23%. Acute Pain Services should aim for incidences less than this standard of care.