OBJECTIVES: Studies have shown that statins may induce diabetes in non-diabetic cardiovascular disease patients, as class effect. It is uncertain if any particular statin is more likely to have greater risk of developing diabetes; only direct and indirect comparison of treatment effects may clarify this uncertainty. We had previously conducted a systematic literature review to estimate the relative likelihood of statin vs. placebo or other agents. This network meta-analysis (NMA) was conducted to rank the treatments according to their risk of developing diabetes. METHODS: We searched databases like Embase, PubMed, and Cochrane. Randomized controlled trials that studied one of the statins and reported incidence of diabetes as an outcome were considered for inclusion. Data synthesis was performed by pairwise meta-analysis and NMA using STATA^® with routines available from www.mtm.uoi.gr