AIMS: The purpose of this study was to determine the effects of cancer occurrence on persistence of oral antidiabetic drugs (OAD) in France. METHODS: A retrospective cohort including incident OAD users between 2006 and 2011 was set up using a permanent sample of health insurance beneficiaries (Echantillon Generaliste de Beneficiaires, EGB). A Cox model was used to assess the association between cancer occurrence and OAD persistence. Non-persistence was defined as a gap in OAD treatment coverage between the end of a given prescription and a new one greater than or equal to 90 days. Cancer occurrence was studied as a time-dependent variable. RESULTS: The study included 13,943 OAD users. Median follow-up was 760 days. After adjustment for age, sex, first OAD used, type of prescriber and polypharmacy, non-persistence risk was higher after a diagnosis of cancer: (HR: 1.93 and IC 95% 1.69; 2.21). Subgroup analyses according to cancer localization found a higher risk of non-persistence for lung cancer (HR: 2.66 and IC 95% 1.68; 4.23) and colorectal cancer (HR: 2.02 and IC 95% 1.40; 2.91). CONCLUSIONS: Our findings indicate there is an association between cancer diagnosis and OAD non-persistence. Additional studies of this type would be useful to evaluate the association between cancer diagnosis and persistence of treatment of other chronic diseases.Read less <