OBJECTIVE: The goal was to review systematically the comparative effectiveness of yoga, compared with other exercise interventions, for older adults as shown on measures of health and physical functioning. DESIGN: This was a systematic review with both narrative synthesis and meta-analysis. DATA SOURCES: Searches were conducted in MEDLINE®/PUBMED, PSYCINFO, CINAHL, Web of Science, and SCOPUS; bibliographies of selected articles; and one systematic review on the effects of yoga on cardiovascular disease. METHODS: Original studies from 1950 to November 2010 were sought, evaluating the effects of yoga on older adults. The search was restricted to randomized controlled trials of yoga in subjects ≥age 60, and published in English. Data were extracted and evaluated regarding setting, population size and characteristics, intervention type and duration, comparison group, outcome assessment, data analysis, follow-up, key results, and the quality of each study according to specific predetermined criteria. RESULTS: The search yielded 18 eligible studies (N=649). The studies reported on older adults across a range of settings, intervention intensity, and outcome measures. The majority of the studies had<35 participants (range 9-77). Quantitative and qualitative synthesis of the studies suggested that the benefits of yoga may exceed those of conventional exercise interventions for self-rated health status, aerobic fitness, and strength. However, the effect sizes were modest, and the evidence was mixed for yoga's effect on depression, sleep, and bone-mineral density. Studies did not find an effect on cognition. CONCLUSIONS: Small studies with mixed methodological quality suggested that yoga may be superior to conventional physical-activity interventions in elderly people. The precision of the estimates remains low. Larger studies are necessary to define better the intersection of populations, settings, and interventions in which yoga is most beneficial.