Vision impairment is common among older adults and affects dementia screening assessments, which include visually presented items. We developed and validated a version of the Montreal Cognitive Assessment (MoCA) for people with vision impairment that includes all the cognitive domains included in the standard MoCA. Visual components of the MoCA were adapted by developing alternative spoken forms. We used both individual item analysis and item substitution to identify the optimal set of alternative items for inclusion in the Montreal Cognitive Assessment for People With Vision Impairment (MoCA-VI) in place of the original items to maximize sensitivity and specificity for dementia. We evaluated the performance and reliability of the final tool, including adjustments for demographic factors. One hundred twenty-eight participants with vision impairment (presenting distance visual acuity worse than 6/12), 79 cognitively healthy and 49 with dementia, completed the adapted MoCA. An additional 86 participants with normal vision completed the standard MoCA and alternative items to assess score equivalence and independence from vision impairment. Twenty-six participants were retested 2-4 weeks after initial testing. With the optimal item set, the final MoCA-VI had an area under the curve of 0.96 (95% CI [0.93, 0.99]). At a cut point of 24 points or less, sensitivity was 95.9%, with a specificity of 92.4%. The intraclass correlation for test-retest reliability was 0.84 (95% CI [0.81, 0.96]). The MoCA-VI is a specific and reliable test for possible dementia among adults with vision impairment. (PsycInfo Database Record (c) 2025 APA, all rights reserved).