A total of 46 patients who had undergone surgery for the relief of temporal lobe epilepsy were tested on the Rivermead Behavioural Memory Test (RBMT) and two traditional tests of memory. The patient and a relative or close friend were also asked to complete a questionnaire about the everyday memory problems experienced by the patients. Scores on the RBMT failed to distinguish between patients who had undergone either a traditional en bloc temporal lobectomy or a more restricted resection, the amygdalo-hippocampectomy, although scores on one of the traditional memory tests did. Epilepsy-related variables did not alter the outcome of data analyses. Only relatives' questionnaire ratings of patients' memory correlated with patients' scores on the RBMT.