Pain not responsive to morphine is often problematic. Animal and clinical studies have suggested that N-methyl-D-aspartate (NMDA) antagonists, such as ketamine, may be effective in improving opioid analgesia in difficult pain syndromes, such as neuropathic pain. A slow bolus of subhypnotic doses of ketamine (0.25 mg/kg or 0.50 mg/kg) was given to 10 cancer patients whose pain was unrelieved by morphine in a randomized, double-blind, crossover, double-dose study. Pain intensity on a 0 to 10 numerical scale; nausea and vomiting, drowsiness, confusion, and dry mouth, using a scale from 0 to 3 (not at all, slight, a lot, awful); Mini-Mental State Examination (MMSE) (0-30); and arterial pressure were recorded before administration of drugs (T0) and after 30 minutes (T30), 60 minutes (T60), 120 minutes (T120), and 180 minutes (T180). Ketamine, but not saline solution, significantly reduced the pain intensity in almost all the patients at both doses. This effect was more relevant in patients treated with higher doses. Hallucinations occurred in 4 patients, and an unpleasant sensation ("empty head") was also reported by 2 patients. These episodes reversed after the administration of diazepam 1 mg intravenously. Significant increases in drowsiness were reported in patients treated with ketamine in both groups and were more marked with ketamine 0.50 mg/kg. A significant difference in MMSE was observed at T30 in patients who received 0.50 mg/kg of ketamine. Ketamine can improve morphine analgesia in difficult pain syndromes, such as neuropathic pain. However, the occurrence of central adverse effects should be taken into account, especially when using higher doses. This observation should be tested in studies of prolonged ketamine administration.