Cystatin C has recently been shown to be associated with incident type 2 diabetes. This study aims to validate this association and to study the impact of baseline adiposity. Methods. We investigated the 3-year diabetes incidence in 2849 participants from the French Data of an Epidemiological Study on the Insulin Resistance syndrome study, without overt kidney disease. Odds ratios (ORs) associated with cystatin C were adjusted for classical diabetes risk factors and interactions between cystatin C and these risk factors were studied. Results. Baseline serum cystatin C was significantly associated with incident diabetes on univariate analysis (OR/1 SD of log cystatin C: 1.74; 95 confidence interval [CI] 1.33-2.28; P = 0.0001) and after adjustment for age and gender (OR 1.55; 95 CI 1.15-2.10; P = 0.0039). This association was independent of serum creatinine-derived measures of baseline renal function and independent of fasting plasma glucose and HbA1c. When body mass index (BMI), waist circumference or baseline insulin resistance index were used as covariates, there was an interaction with cystatin C level. Cystatin C was associated only with incident diabetes for people with BMI, waist circumference or insulin resistance index = median value with OR (95 CIs), respectively: 1.35 (0.98-1.84, P = 0.0625); 1.39 (1.01-1.91, P = 0.0441) and 1.41 (1.02-1.94, P = 0.0398). Conclusions. Cystatin C was associated with 3-year incident diabetes but only in people with central adiposity or insulin resistance. This should be considered in further studies assessing the clinical relevance of its prognostic value.