Surprisingly few ecologically-valid assessments of executive function exist, but the Baycrest Multiple Errands Test (BMET) shows promise in identifying executive impairment. The goal of the present study was to develop both a revised version of the assessment (BMET-R), to improve the test's ability to discriminate between patients and healthy participants, and an alternate form of the BMET-R to permit repeat testing. Sixteen individuals with acquired brain injury (ABI) due to stroke or trauma and 16 healthy participants completed a series of neuropsychological assessments, questionnaires, the BMET-R and its alternate form (in counterbalanced order). The results of the study indicated that participants with ABI omitted more tasks, broke more rules, and were less efficient than healthy participants on both the revised BMET-R and its alternate form. Moreover, significant correlations were found between the two versions of the BMET-R for task completions, omissions, errors, rule breaks and inefficiencies but few significant correlations were observed between the BMET-R versions and measures of executive dysfunction in everyday life. These results indicate that the two versions of the BMET-R are able to dissociate the performance of participants with ABI from that of healthy participants. However, despite overlaps in performance and correlations between the two versions of the BMET-R, they did not identically assess executive deficits. This suggests that caution should be used when constructing and validating alternate versions of performance-based assessments. [ABSTRACT FROM AUTHOR]