This study was aimed at the assessment of incidence of malignancies in type 2 diabetic patients treated with different sulphonylureas. A matched case-control study was performed. Cases were 195 diabetic patients aged 69.0 +/- 9.2 years who had an incident malignancy. Controls were 195 diabetic patients, unaffected by cancer, who were matched with the corresponding case for age, sex, duration of diabetes, BMI, HbA1(c), comorbidity, smoking and alcohol abuse. Exposure to hypoglycaemic drugs during the 10 years preceding the event (or matching index date) was assessed. After adjusting for concomitant therapies, exposure to metformin and gliclazide for more than 36 months was associated with a significant reduction in the risk of cancer (adj. ORs with 95% CI: 0.28 (0.13-0.57), p < 0.001, and 0.40 (0.21-0.57), p = 0.004, respectively). Conversely, use of glibenclamide for at least 36 months was associated with increased incidence of malignancies (adj. OR 2.62 (1.26-5.42); p = 0.009). Treatment with insulin, thiazolidinediones, or acarbose, was not associated with significant differences in the incidence of cancer. Long-term treatments with individual sulphonylureas could have differential effects on the risk of cancer. In particular, the possible protective effect of gliclazide, as well as the risk associated with glibenclamide, deserves further investigation.