The purpose of this metaanalysis was to determine the benefits of postoperative epidural analgesia in patients operated on under general anesthesia. By searching the American National Library of Medicine's Pubmed database from 1966 to July 10, 2004, 70 studies were identified. These included 5402 patients, of which 2660 had had epidural analgesia. Epidural analgesia reduces the incidence of arrhythmia, odds ratio (OR) = 0.59 (95%CI = 0.42, 0.81, P = 0.001); time to tracheal extubation, OR = -3.90 h (95%CI = -6.37, -1.42, P = 0.002); intensive care unit stay, OR = -2.94 h (95%CI = -5.66, -0.22, P = 0.03); visual analogical pain (VAS) scores at rest, OR = -0.78 (95%CI = -0.99, -0.57, P < 0.00001) and during movement, OR = -1.28 (95%CI = -1.81, -0.75, P < 0.00001); maximal blood epinephrine, OR = -165.70 pg.ml(-1) (95%CI = -252.18, -79.23, P = 0.0002); norepinephrine, OR = -134.24 pg.ml(-1) (95%CI = -247.92, -20.57, P = 0.02); cortisol, OR = -55.81 nmol.l(-1) (95%CI = -79.28, -32.34, P < 0.00001); and glucose concentrations achieved, OR = -0.87 nmol.l(-1) (95%CI = -1.37, -0.37, P = 0.0006). It also reduces the first 24-h morphine consumption, OR = -13.62 mg (95%CI = -22.70, -4.54, P = 0.003), and improves the forced vital capacity (FVC), OR = 0.23 l (95%CI = 0.09, 0.37, P = 0.001) at 24 h. A thoracic epidural containing a local anesthetic reduces the incidence of renal failure: OR = 0.34 (95%CI = 0.14, 0.81, P = 0.01). Epidural analgesia may thus offer many advantages over other modes of postoperative analgesia.