We investigated whether intraoperative 'subanesthetic doses' of ketamine have a postoperative anti-hyperalgesic and an analgesic effect and which is the preferential route of administration, either systemic (intravenous, i.v.) or epidural. One hundred patients scheduled for rectal adenocarcinoma surgery under combined epidural/general anesthesia were included. Before skin incision all the patients received an epidural bolus followed by an infusion of continuous bupivacaine/sufentanil/clonidine mixture. They were randomly assigned to receive no ketamine (group 1), i.v. ketamine at the bolus dose of 0.25 mg/kg followed by an infusion of 0.125 mg/kg per h (group 2), 0.5 mg/kg and 0.25 mg/kg per h (group 3), epidural ketamine 0.25 mg/kg and 0.125 mg/kg per h (group 4), or 0.5 mg/kg and 0.25 mg/kg per h (group 5). All i.v. and epidural analgesics were stopped at the end of surgery and patients were connected to an i.v. morphine patient-controlled analgesia (PCA) device. Short-term postoperative analgesia (72 h) was assessed by pain visual analog scale scores at rest, cough, and movements as well as by PCA requirements. Wound mechanical hyperalgesia was evaluated and residual pain was assessed by asking the patients at 2 weeks, and 1, 6, and 12 months. The area of hyperalgesia and morphine PCA requirements were significantly reduced in group 3. These patients reported significantly less residual pain until the sixth postoperative month. These observations support the theory that subanesthetic doses of i.v. ketamine (0.5 mg/kg bolus followed by 0.25 mg/kg per h) given during anesthesia reduce wound hyperalgesia and are a useful adjuvant in perioperative balanced analgesia. Moreover, they show that the systemic route clearly is the preferential route.