Aim: Different intervention strategies can prevent type 2 diabetes (T2DM). Aim of the present systematic review and meta-analysis was to evaluate the effectiveness of different strategies. Methods: Studies were grouped into 15 different strategies: 1: diet plus physical activity; 2: physical activity; 3-6: anti-diabetic drugs [glitazones, metformin, beta-cell stimulating drugs (sulphanylureas, glinides), alfa-glucosidase inhibitors]; 7-8: cardiovascular drugs (ACE inhibitors, ARB, calcium antagonists); 9-14 [diets, lipid-affecting drugs (orlistat, bezafibrate), vitamins, micronutrients, estrogens, alcohol, coffee]; 15: bariatric surgery. Only controlled studies were included in the analysis, whether randomized, non-randomized, observational studies, whether primarily designed to assess incident cases of diabetes, or performed with other purposes, such as control of hypertension, of ischemic heart disease or prevention of cardiovascular events. Appropriate methodology [preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement] was used. Seventy-one studies (490 813 subjects), published as full papers, were analysed to identify predictors of new cases of T2DM, and were included in a meta-analysis (random-effects model) to study the effect of different strategies. Intervention effect (new cases of diabetes) was expressed as odds ratio (OR), with 95% confidence intervals (C.I.s). Publication bias was formally assessed. Results: Body mass index was in the overweight range for 13 groups, obese or morbidly obese in lipid-affecting drugs and in bariatric surgery. Non-surgical strategies, except for beta-cell stimulating drugs, estrogens and vitamins, were able to prevent T2DM, with different effectiveness, from 0.37 (C.I. 0.26-0.52) to 0.85 (C.I. 0.77-0.93); the most effective strategy was bariatric surgery in morbidly obese subjects [0.16 (C.I. 0.11,0.24)]. At meta-regression analysis, age of subjects and amount of weight lost were associated with effectiveness of intervention. Conclusions: These data indicate that several strategies prevent T2DM, making it possible to make a choice for the individual subject.