OBJECTIVE: Discussions about statin efficacy in cardiovascular prevention are always based on data from blinded randomized controlled trials (RCTs) comparing statin to placebo; however, discussion of side effects is not. Clinicians often assume symptoms occurring with statins are caused by statins, encouraging discontinuation. We test this assumption and calculate an evidence-based estimate of the probability of a symptom being genuinely attributable to the statin itself. METHODS: We identified RCTs comparing statin to placebo for cardiovascular prevention that reported side effects separately in the two arms. RESULTS: Among 14 primary prevention trials (46,262 participants), statin therapy increased diabetes by absolute risk of 0.5% (95% CI 0.1-1%, p = 0.012), meanwhile reducing death by a similar extent: -0.5% (-0.9 to -0.2%, p = 0.003). In the 15 secondary prevention RCTs (37,618 participants), statins decreased death by 1.4% (-2.1 to -0.7%, p < 0.001). There were no other statin-attributable symptoms, although asymptomatic liver transaminase elevation was 0.4% more frequent with statins across all trials. Serious adverse events and withdrawals were similar in both arms. CONCLUSIONS: Only a small minority of symptoms reported on statins are genuinely due to the statins: almost all would occur just as frequently on placebo. Only development of new-onset diabetes mellitus was significantly higher on statins than placebo; nevertheless only 1 in 5 of new cases were actually caused by statins. Higher statin doses produce a detectable effect, but even still the proportion attributable to statins is variable: for asymptomatic liver enzyme elevation, the majority are attributable to the higher dose; in contrast for muscle aches, the majority are not.