We describe the target-controlled administration of propofol and remifentanil, combined with monitoring of the bispectral index, during an awake craniotomy for removal of a left temporo-parietal tumour near the motor speech centre. Target concentrations of the two drugs were adjusted according to the patient's responses to painful stimuli and surgical events, and the need for speech testing. Allowing the effect-site concentrations of propofol and remifentanil to decrease during surgery allowed the performance of cortical speech mapping and the testing of the patient's ability to speak. Although the bispectral index was not used as a guide for the administration of the drugs, its value correlated better with the patient's responsiveness than did the predicted effect-site concentrations of propofol. Side-effects, comprising hypotension, respiratory depression and airway obstruction, were related to rapid increases in drug infusion rates and were easily managed.