OBJECTIVES: Low adherence to medications in chronic diseases is a well documented problem that diminishes drug actions. Understanding the factors that can impact adherence is a priority in order to improve it. Diabetes adherence can be influenced by the complexity of administration of some medications as well as patients’ attitudes and doctor/patients interaction. This study evaluates the influence of the following variables in the adherence of people with diabetes: Dosage, patient age, gender, substance, price of drug, pack size, pack strength and doctor specialty. METHODS: Adherence was calculated using the medical possession ratio (MPR) = unique days of therapy/available days. A logistic regression was conducted to explore the multiple variables affecting the MPR. RESULTS: The analyses included 25 months (from July 2009 to July 2011) of longitudinal prescription data for Germany for the ATC class A10 (oral antidiabetic). We found significant effects on adherence for age groups 55-64 (0.0909 p<0.01), 65-74 (0.1266, p<0.0004) and > 75 (0.0868, p<0.0155), dosage (-0.7646, p<0.0001), gender (-0.0493 p<0.001), pack size (0.0119, p <.0001), doctor specialty (0.0251, p<0.0286) and 78% of all substances.   CONCLUSIONS: Adherence seems to be negatively influence by dosage and younger age groups.  Price does not seem to have an influence. Perhaps this is because the healthcare system in Germany reimbursed all oral antidiabetics. Metformin has the greater odds for poor adherence aside from the substances “glitazones”. Lower adherence rate might be negatively influenced by the recent warnings for Glitazones. Gender is a weak predictor for adherence. When comparing specialists with general practitioners, the odds of adherence are not very different between them (slightly positive towards the specialist group_ coeff 0.0251, p<0.0286).  This could be related to the diabetes management plan that exists in Germany since 2003, which has several aspects, from support to clinicians to training of patients.