A substantial number of brain injured patients complain of memory deficits, despite achieving scores within the normal range on tests of memory functioning. The Rivermead Behavioural Memory Test is an ecologically valid test used to assess everyday memory problems. This test is effective at detecting moderate to-severe impairments, yet subtle memory deficits may go undetected for some patients who achieve a score within the normal range. The initial development of an extended version of this test (RBMT-E), designed to detect subtle decrements in memory performance, has recently been described. The performance of 16 brain injured patients was assessed on both the RBMT and the KBMT-E. The performance of these patients on the RBMT-E was compared with that of matched controls. Overall, the patients performed significantly worse than the controls, and showed particular difficulty in two subtests involving recalling a route and remembering to deliver a message. Those patients who scored in the 'normal' range on the RBMT could be further differentiated on the basis of their RBMT-E scores into 'good', 'average' and 'poor' performance categories. The patients' performance was not significantly associated with general intellectual ability. These results suggest that the RBMT-E may be a useful clinical tool to aid therapists in the assessment of subtle impairments of everyday memory performance following brain injury.