We investigated the association between specific types of physical activity and the risk of type 2 diabetes in a systematic review and meta-analysis of published studies. PubMed, Embase and Ovid databases were searched for prospective studies and randomized trials up to 2nd of March 2015. Summary relative risks (RRs) were calculated using a random effects model. Eighty-one studies were included. The summary RRs for high versus low activity were 0.65 (95 % CI 0.59–0.71, I2 = 18 %, n = 14) for total physical activity, 0.74 (95 % CI 0.70–0.79, I2 = 84 %, n = 55) for leisure-time activity, 0.61 (95 % CI 0.51–0.74, I2 = 73 %, n = 8) for vigorous activity, 0.68 (95 % CI 0.52–0.90, I2 = 93 %, n = 5) for moderate activity, 0.66 (95 % CI 0.47–0.94, I2 = 47 %, n = 4) for low intensity activity, and 0.85 (95 % CI 0.79–0.91, I2 = 0 %, n = 7) for walking. Inverse associations were also observed for increasing activity over time, resistance exercise, occupational activity and for cardiorespiratory fitness. Nonlinear relations were observed for leisure-time activity, vigorous activity, walking and resistance exercise (p nonlinearity < 0.0001 for all), with steeper reductions in type 2 diabetes risk at low activity levels than high activity levels. This meta-analysis provides strong evidence for an inverse association between physical activity and risk of type 2 diabetes, which may partly be mediated by reduced adiposity. All subtypes of physical activity appear to be beneficial. Reductions in risk are observed up to 5–7 h of leisure-time, vigorous or low intensity physical activity per week, but further reductions cannot be excluded beyond this range.