The long-term effects of bariatric surgery have yet to be established, and a number of important studies have recently emerged. This systematic review aimed to assess the effects of bariatric surgery on all-cause mortality, cardiovascular events, and cancer compared to non-surgical treatment. We searched PubMed, EMBASE, and CENTRAL up to July 13, 2015, and included randomized controlled trials (RCTs) and non-randomized controlled studies comparing bariatric surgery versus non-surgical treatment and reporting data on the three defined outcomes at 1 year or longer. We analyzed RCTs and non-randomized controlled studies, respectively. Eleven RCTs, 4 non-randomized controlled trials, and 17 cohort studies were included. The randomized evidence suggested substantial uncertainty regarding the effects on all-cause mortality (0/382 vs. 1/287; 7 studies), cancer (OR 0.77, 95 % CI 0.22 to 2.71; 4 studies), and cardiovascular events (no data). The pooled adjusted estimates from non-randomized studies suggested that, compared to the control, the surgical group had lower risk of all-cause mortality (OR 0.55, 95 % CI 0.46 to 0.65; 10 studies), cancer (OR 0.74, 95 % CI 0.65 to 0.85; 2 studies), and cardiovascular events (MI: OR 0.71, 95 % CI 0.54 to 0.94; stroke: OR 0.66, 95 % CI 0.49 to 0.89; and their composite: OR 0.67, 95 % CI 0.54 to 0.83; 1 study). In conclusion, bariatric surgery could reduce all-cause mortality and probably reduce the risk of any type of cancer. The inference was, however, based on studies with limited methodological rigor. Uncertainty remains regarding the effects on cardiovascular events.