The effectiveness of diclofenac 50 mg t.i.d. as additive treatment to parenteral patient-controlled administration therapy (PCAT) with morphine in cancer pain has been investigated in a double-blind study. In the fifteen patients who completed the study, morphine i.v. was titrated to optimal pain relief over 5 days. The mean total morphine consumption was significantly reduced during diclofenac administration (82.8 mg morphine per day) compared to placebo (95.0 mg morphine per day). The reduction in mean morphine consumption during active treatment with diclofenac was independent of the initial dose of self-titrated morphine. Pain, self-assessed according to VAS, tended to be lower during the diclofenac period, although the difference did not reach statistical significance. No adverse events were recorded among the 15 patients who completed the study. The present findings show that a non-steroidal anti-inflammatory agent, such as diclofenac, has a morphine-sparing effect in morphine-treated patients with cancer pain.