Eighteen inpatients receiving morphine for cancer pain in a palliative care unit were recruited to a study employing a range of neuropsychological tests to assess cognitive function. The tests employed were National Adult Reading Test, Williams Delayed Recall Test, Immediate Memory for Digits, Trailing Making Test, and the Digit Symbol Substitution Test. These data were correlated with biochemical tests of renal and hepatic function, morphine dose, route of administration, plasma morphine, morphine-3-glucuronide (M3G) and morphine-6-glucuronide (M6G) concentrations. Despite having no clinical evidence of impairment of cognitive function, the level of current intellectual functioning (Symbol Digit Substitution Test) was on average two standard deviations below normal. Immediate memory appeared to be well preserved, but Delayed Recall and Trailing Making Test scores were significantly above normal. There was no significant correlation between morphine dose, or plasma morphine and M3G and M6G concentrations, and the neuropsychological test results, although a weak correlation was found between plasma morphine concentration and digits forward (r = -0.47, p < 0.05) and Digit Symbol Substitution scores (r = -0.46, p < 0.05). Seven patients had some degree of nausea or vomiting, ascribed as an opioid adverse effect, and had higher serum creatinine concentrations, worse neuropsychological performance, and significantly higher plasma M3G concentrations (p < 0.05). These data provide some evidence to suggest that cognitive functioning in patients with advanced cancer receiving morphine may be significantly impaired despite apparent clinical normality. From these data it is not possible to determine what relative causal contribution the disease and the drug made to these observations, although renal function, plasma morphine, and M3G concentrations may be important. Future research should address a broad range of neuropsychological testing to assist in the modification of practices aimed at enhancement of quality of life, such as opioid substitution or rotation.