Because individual variation is a likely factor affecting both the incidence and severity of side effects and the analgesic efficacy of epidural opioids, assessment of individual variation could be useful in deciding optimal dosage. By evaluating the response to a small test dose of IV fentanyl, we designed this study to predict the degree of pain relief and the incidence of side effects in patients who would be receiving postoperative epidural fentanyl. Before the induction of anesthesia, 50 mug of fentanyl was administered IV, and 2 min after fentanyl, the patient response was evaluated. Twenty-three patients, who reported nausea, sleepiness, dizziness, sensation of warmth, and other symptoms, were categorized as responders (Group R); the remaining 20 patients were categorized as non-responders (Group NR). At the completion of surgery, infusion of epidural fentanyl was administered (0.3 mg/d in 0.25% bupivacaine) for 96 h. At postoperative Hours 6 and 24, Group R had significantly lower visual analog scale scores for postoperative pain intensity and required fewer analgesics than Group NR. The incidence of side effects, however, was 74% for Group R and 10% for Group NR (P < 0.05), and side effects were more serious in Group R. This study demonstrates that preoperative administration of a small dose of fentanyl during the induction of anesthesia enables prediction of the analgesic efficacy of postoperative epidural fentanyl and the incidence and severity of side effects.