We conducted a systematic review of the association between HMG-CoA reductase inhibitor (statin) use and cancer risk. We searched MEDLINE, EMBASE, Web of Science, ISI Proceedings and BIOSIS Previews bibliographic databases, electronic trials registers and reference lists for potentially eligible randomized trials and observational studies. Thirty-eight individual studies (26 randomized trials involving 103,573 participants and 12 observational studies with 826,854 participants) were included. Median follow-up was 3.6 and 6.2 years for trials and observational studies, respectively. In metaanalyses of randomized trials, there was no evidence that statin therapy was associated with incidence of all-cancers (26 trials; pooled risk ratio = 1.00; 95% CI 0.95–1.05;I2 = 0%) or the following site-specific cancers: breast (7 trials; risk ratio = 1.01; 0.79–1.30;I2 = 43%), prostate (4 trials; risk ratio = 1.00; 0.85–1.17;I2 = 0%), colorectum (9 trials; risk ratio = 1.02; 0.89–1.16;I2 = 0%), lung (9 trials; risk ratio = 0.96; 0.84–1.09;I2 = 0%), genito-urinary (5 trials; risk ratio = 0.95; 0.83–1.09;I2 = 0%), melanoma (4 trials; risk ratio = 0.86; 0.62–1.20;I2 = 17%) or gastric (1 trial; risk ratio = 1.00; 0.35–2.85). There was no evidence of differential effects by length of follow-up, statin type (lipophilicvs. lipophobic) or potency. Trial results were generally consistent with observational studies. We conclude that statin use is not associated with short-term cancer risk, but longer-latency effects remain possible. © 2006 Wiley-Liss, Inc.