Objective: Performance-based evaluation of executive function by using real-world daily living activities is an important area of study. This approach has been used extensively in evaluating patients after stroke or traumatic brain injury and patients with schizophrenia. Most important is the fact that until now, there has been no validated performance-based evaluation of executive function in people with dementia. Methods: To address that knowledge gap, this study recruited 80 patients diagnosed with dementia and 80 demographically matched healthy controls. The participants were administered tests for evaluating their performance-based executive function (Chinese Multiple Errands Test), their instrumental activities of daily living (Lawton Instrumental Activities of Daily Living Scale, Chinese Version), and their functional disability (Chinese Version of the Disability Assessment for Dementia), along with a cognitive screening test (Montreal Cognitive Assessment, Hong Kong Version) and a neuropsychological test of executive function (Trail-making Test). Results: The Chinese Multiple Errands Test demonstrated excellent inter-rater reliability, test-retest reliability and high internal consistency. Results revealed that the healthy controls out-performed the dementia patients in the performance-based executive function and cognitive screening, but not in the instrumental activities of daily living tests. Additionally, the performance efficiency scores of the older adults with dementia on the Chinese Multiple Errands Test correlated significantly with their performance results on the neuropsychological test of executive function and on the tests of functional disability and cognitive function. Conclusion: Our results indicated that the Chinese Multiple Errands Test is a reliable and valid instrument for assessing executive function in Chinese older people with dementia.