IntroductionDementia is a progressive brain degeneration characterized by a progressive deterioration in cognition and independent living capacity. Since dementia is a complex syndrome, multimodal non-pharmacological interventions (MNPIs) are highly recommended. Currently, there is less available evidence to describe the content, length, and frequency of multimodal interventions for cognitive function improvement for people with dementia (PWD).MethodA comprehensive search was performed in PubMed, EMBASE, CINAHL, Web of Science, and Medline international databases. The quality appraisal of the studies was done by the Cochrane risk of bias assessment tools.ResultsA total of 19 controlled trial studies were included. Most of the included studies reported that MNPIs resulted in improvement, stability, or attenuation of decline in cognitive function of PWD. The reported effectiveness of MNPIs on cognitive function ranged from medium (0.29 Cohen's d) to large (2.02 Cohen's d) effect sizes. The median duration of intervention was 12 weeks for a 1-h session.ConclusionThis systematic review showed that MNPIs might improve people's cognitive functions for PWD. Physical exercise, music, and cognitive interventions were used in the content of multimodal interventions in a majority of the studies. Therefore, high-quality randomized controlled trial (RCT) studies with repeated-measured design on the combined effect of physical exercise, music, and cognitive intervention on cognitive function for PWD are recommended.Systematic Review Registrationhttp://www.crd.york.ac.uk/PROSPERO/, identifier: CRD42020222065.