Background and objectives - In France, diabetes mellitus is now the second cause of end stage renal disease. In a large previous French national study, we observed that dialyzed diabetics have a significant lower risk of death by cancer. This first study was focused on cancer death but did not investigate cancer incidence. In this context, the aim of this second study was to compare the incidence of cancer in diabetic dialyzed patients compared to non-diabetic dialyzed patients in a French region. Methods - This epidemiologic multicentric study included 588 diabetic and non-diabetic patients starting hemodialysis between 2002 and 2007 in Bretagne. Data were issued from REIN registry and cancer incidence were individually collected from medical records. Diabetics and non-diabetics were matched one by one on age, sex and year of dialysis initiation. Results - During the follow-up, we observed 28 cancers (9.4%) in diabetic patients and 26 cancers (8.9%) in non-diabetics patients. The cumulative incidence to develop a cancer 2 years after the dialysis start was approximately 6% in both diabetics and non-diabetics patients. In univariate Fine and Gray analysis, BMI, hemoglobin, statin use had P-value<0.2. However, in the adjusted model, these variables were not significantly associated with cancer incidence. Conclusion - This study lead on a little number of dialyzed patients did not show any significant difference on cancer incidence between diabetic and non-diabetic patients after hemodialysis start.