OBJECTIVE To examine discriminant validity and test-retest reliability of the Zürich maxi mental status inventory (ZüMAX) in patients with stroke. BACKGROUND The ZüMAX is a novel domain-specific cognitive assessment tool to screen for disturbances in neuropsychological function. The test can be used in stroke rehabilitation to estimate severity of cognitive impairment. Because evidence for validity and reliability is lacking, the tool's clinical use is limited. METHODS We administered the ZüMAX in a test-retest design to 33 community-dwelling stroke survivors, and once to 35 healthy controls matched for age and sex. RESULTS We found significant group differences in subscores for the cognitive domains of executive functions and language as well as total score (P=0.001 to 0.004); we did not find group differences for the domains of praxia (defined as the ability to perform purposeful actions), visual perception and construction, or learning and memory. Test-retest reliability of the total score was good (intraclass correlation coefficient=0.81), with the individual domain subscores ranging from poor to fair (0.59 to 0.79). The ZüMAX could detect changes in patients with low smallest detectable differences in executive functions, language, and praxia (0.05 to 1.49) and total score (0.09). CONCLUSION The ZüMAX has moderate to good test-retest reliability. Furthermore, the tool might discriminate between healthy persons and chronic stroke survivors on three of five subscales. The ZüMAX shows promise in measuring neuropsychological disturbances in stroke survivors; however, further trials are required with larger samples.