Purpose Women with polycystic ovary syndrome (PCOS) exhibit many metabolic abnormalities that are associated with an increased cardiovascular disease risk. Exercise may promote improvements in lipid profile and insulin sensitivity in women with PCOS. There is however, a knowledge gap on the optimal dose of exercise, regarding duration, intensity, type, and frequency of exercise. The aim of this systematic review and meta-analysis was to define effective types of exercise to improve cardiometabolic profile in PCOS. Methods We included randomised controlled trials (RCT), quasi-RCT, and controlled clinical trials focusing on reproductiveaged women diagnosed with PCOS. Eligible interventions included those with at least two weeks of supervised exercise sessions. Primary outcomes were blood lipids, blood glucose, blood pressure, measures of abdominal adiposity, and inflammation markers. Secondary outcomes were total and free testosterone, sex hormone binding globulin, and measures of insulin resistance. Nine electronic databases were searched from inception to present for English language publications. The Cochrane Risk Assessment tool was used to assess bias in the included studies. Outcomes were quantitatively synthesised and a meta-analysis was performed. Pooled effect estimates and 95% confidence intervals were presented. Results This systematic review identified three trials, including 231 participants with PCOS, that examined the effect of structured, supervised exercise on cardiometabolic outcomes. Analysis of pooled data indicated statistical favourable effects of exercise on total cholesterol, fasting glucose, waist circumference and waist-to-hip ratio, systolic blood pressure, C-reactive protein, total testosterone, and sex hormone binding globulin using post-intervention scores. Conclusions Moderate aerobic exercise interventions ≥3 months in duration, with a frequency of 3/week for at least 30-min, may have favourable effects on various cardiometabolic risk factors in women with PCOS. However, results should be interpreted with caution. Many of the outcomes were based on studies with serious methodological limitations, and only one "gold-standard" RCT was identified. PROSPERO ID: CRD42018086117.