Although patient-controlled analgesia (PCA) pumps have been in use for more than a decade, the optimal PCA analgesic has yet to be identified. Many drugs are used; however, morphine remains the "gold standard" of opioid analgesics worldwide. The present study evaluated morphine and hydromorphone (Dilaudid) PCA with respect to analgesic efficacy, side effects, mood, and cognitive function. Sixty-one opioid naive patients undergoing lower abdominal surgery participated in the double-blind protocol. Verbal rating scores, use of medication, and side effects for the two medications were recorded. Cognitive functioning was assessed by computation of Digit Symbol and Trails Making B Tests. Self-reported affective state (mood) was measured by Profile of Mood States (POMS) inventory. Both medications provided adequate analgesia without a difference in side effects. Cognitive performance was poorer in the hydromorphone group (P < 0.05). Patients receiving hydromorphone reported less anger/hostility (P < 0.01) and generally better mood elevations on the other subscales than those receiving morphine. A similar incidence of side effects and dose medication can be anticipated with morphine and hydromorphone. When considering cognitive effects, morphine had less adverse consequences, while hydromorphone appeared to result in improved mood. We conclude that hydromorphone may provide a suitable alternative to morphine.