The analgesic efficacy and adverse effects of morphine and oxymorphone in 32 patients who received traditional patient-controlled analgesia (PCA) following cesarean delivery were compared with those in 32 other patients receiving the same agents via PCA plus basal opioid infusion (PCA + BI). All patients were operated upon during epidural anesthesia with 2% lidocaine and 1:200,000 epinephrine to achieve a T4 sensory level. Upon first complaint of pain in the recovery room, patients were given a titrated iv loading dose of the assigned opioid until comfortable and were then provided with a programmable PCA device. Group I (PCA) consisted of two subsets in which incremental boluses of morphine (1.8 mg, n = 16) or oxymorphone (0.3 mg, n = 16) could be self-administered via conventional PCA. Patients in group II (PCA + BI) received a basal infusion of morphine (0.6 mg/hour, n = 16) or oxymorphone (0.1 mg/hour, n = 16) in addition to self-administered boluses of 1.8 and 0.3 mg, respectively. Patients were evaluated for 24 h following initiation of analgesic therapy, and 10-cm visual analog scales (VAS) were utilized at selected intervals to assess pain at rest, pain during movement, and satisfaction with therapy. The level of sedation and incidence of nausea/vomiting and pruritus were also recorded. Patients utilizing PCA + BI noted significant reductions in resting pain scores with oxymorphone and decreased pain during movement with both opioids when compared with individuals using PCA alone (P less than 0.05). There were no significant differences between treatment groups in 24-h dose requirements or patient satisfaction with therapy (P = ns).(ABSTRACT TRUNCATED AT 250 WORDS)