Older adults with cognitive impairment have deficits in executive systems that affect their gait automaticity. The aim of the meta-analysis and systematic review was to examine the effects of interventions focus on only motor learning principles on gait performance and physical functions (e.g., dynamic balance). After inspections of 879 articles, 11 relevant studies were selected for systematic review and meta-analysis. The PEDro scale and Modified Downs and Black checklist were used to assess the quality of studies, and a random-effect model was used at a 95% confidence interval for calculating pooled effect sizes. The results of this systematic review and meta-analysis showed that motor learning interventions increased gait speed, cadence, stride length, and reduced gait cognitive cost but did not affect gait variability and physical function. In conclusion, practitioners should pay attention more to the potential benefits of motor learning interventions in rehabilitating older adults with cognitive impairment.