Objective There is a long‐standing debate in the literature whether depressive symptoms increase the risk of dementia in older with mild cognitive impairment (MCI). We aim to conduct a meta‐analysis of studies that evaluated the risk of dementia in subjects with MCI and depressive symptoms compared with subjects with MCI and no depressive symptoms. Methods We calculated the relative risk of progression to dementia in subjects with MCI and depressive symptoms compared with subjects with MCI and no depressive symptoms using a generic inverse variance method with random effect models. Results Eighteen studies were included in the meta‐analysis, with a sample size of 10,861 MCI subjects. The pooled relative risk of progressing to dementia was 1.28 CI95% [1.09–1.52] (p = 0.003) in the group of MCI subjects with depressive symptoms compared with the MCI subjects with no depressive symptoms. Discussion Our results provide additional evidence that depressive symptoms determine an additive risk effect to the progression to dementia in subjects with MCI. The comorbidity between depression and cognitive impairment can be an intervention target for prevention of dementia in MCI subjects. Copyright © 2015 John Wiley & Sons, Ltd.