Background: Research in both humans and animals indicates that physical activity can enhance cognitive activity, but whether this is true in patients with stroke is largely unknown.We aimed to evaluate the relationship between increased physical activity after stroke and cognitive performance. Methods: A systematic review was conducted of MEDLINE, EMBASE, PsycINFO and other electronic databases. All randomized controlled trials and controlled clinical studies that evaluated the effect of physical activity or exercise on cognitive function in stroke were included. Study quality was assessed using four criteria concerning sources of bias (use of randomization, allocation concealment, blinding of outcome assessment, whether all patients were accounted for in outcome data). Results: The literature search (first run in 2008, updated in 2011) yielded 12 studies that satisfied inclusion criteria. Exercise interventions were heterogeneous; some studies compared different intensities of movement rehabilitation, others included a specific exercise program. Cognitive function was rarely the primary outcome measure, and cognitive assessment tools used were generally suboptimal. Nine studies had sufficient data to be included in a meta-analysis, which indicated a significant benefit of intervention over control (SMD = 0.20, 95% CI: 0.04–0.36; z = 2.43, p = 0.015). Studies that met all four quality criteria reported smaller treatment benefit than studies that did not. Conclusions: There is some evidence that increased physical activity after stroke enhances cognitive performance. The pool of studies identified, however, was small and methodological shortcomings were widespread.