Forty ASA I-III patients recovering from major abdominal or orthopedic operations were investigated in an open clinical study to evaluate analgesic efficacy and threshold plasma concentrations of alfentanil during the early postoperative period using patient-controlled analgesia (PCA) by means of the On-Demand Analgesia Computer. Alfentanil demand dose was 212 micrograms, continuous infusion rate 25 micrograms/hr, hourly maximum dose 1.5 mg/hr; the lockout time was set to 1 min. The duration of PCA was 18.1 +/- 5.2 hr (mean, SD) during which time 23.8 +/- 14.2 demands per patient were recorded, resulting in an average alfentanil consumption of 4.99 +/- 3.03 micrograms/kg/hr. Patient acceptance of PCA was high. Side effects were only of minor intensity and never gave rise to concern. Based on our own earlier PCA experience with other opiate analgesics, alfentanil proved to be about 1/15th as potent an analgesic as fentanyl and about 6-7 times more potent than morphine if both intensity and duration of effect were considered. Minimum effective alfentanil plasma concentration (MEC) varied greatly and could be best described by a lognormal distribution (range 0.6-99.2 ng/mL, median 14.9 ng/mL). Intraindividual MEC variability was consistently lower than intersubject variability (37.0% vs 65.2%).