Aims To date, there is insufficient evidence available to compare the outcome of cemented and uncemented fixation of the humeral stem in reverse shoulder arthroplasty (RSA). Methods A systemic review comprising 41 clinical studies was performed to compare the functional outcome and rate of complications of cemented and uncemented stems in RSA. These included 1455 cemented and 329 uncemented shoulders. The clinical characteristics of the two groups were similar. Variables were compared using pooled frequency-weighted means and relative risk ratios (RR). Results Uncemented stems had a significantly higher incidence of early humeral stem migration (p < 0.001, RR 18.1, 95% confidence interval (CI) 5.0 to 65.2) and non-progressive radiolucent lines (p < 0.001, RR 2.4, 95% CI 1.7 to 3.4), but a significantly lower incidence of postoperative fractures of the acromion compared with cemented stems (p = 0.004, RR 14.3, 95% CI 0.9 to 232.8). There was no difference in the risk of stem loosening or revision between the groups. The cemented stems had a greater relative risk of infection (RR 3.3, 95% CI 0.8 to 13.7), nerve injury (RR 5.7, 95% CI 0.7 to 41.5) and thromboembolism (RR 3.9, 95% CI 0.2 to 66.6). The functional outcome and range of movement were equivalent in the two groups. Discussion RSA performed with an uncemented stem gives them equivalent functional outcome and a better complication profile than with a cemented stem. The natural history and clinical relevance of early stem migration and radiolucent lines found with uncemented stems requires further long-Term study. Take home message: This study demonstrates that uncemented stems have at least equivalent clinical and radiographic outcomes compared with cemented stems when used for reverse total shoulder arthroplasty.