OBJECTIVES: Diabetes is a chronic condition for which effective medications must be taken continuously (i.e. patient must be observant). Research shows that introducing user fees may reduce patient observance. However, there is little evidence that providing free medications improves diabetic patient observance. Taking advantage of a change in Quebec (Canada) policy the objective of this study was to assess whether the return to free medications (RFM) improved diabetic patient observance. METHODS: This study used a pretest-posttest design: drug use by diabetic patients in the 3 years prior to RFM was compared to their drug use in the 2 years following RFM.  Data came from the Quebec public drug plan (QPDP) for three groups: social welfare recipients, elderly receiving full guaranteed revenue supplement (GRS) and those receiving partial GRS. Data on oral antidiabetic drug consumption were obtained for a random sample of patients that were covered by the QPDP during the full five year period (N total = 3308). Two indicators of observance were measured: whether a patient was using antidiabetic drugs at least 80% of days and the proportion of days where antidiabetic drugs were used. These were measured both for the 12 months following the first prescription prior to RFM and after RFM and for the whole pre-RFM and post-RFM periods. To compare pre and post RFM data, we used Chi-square test for the first indicator and t-test for the second indicator. RESULTS: The proportion of patients who were observant was significantly (p<0.001) higher after RFM compared to before RFM, both for 12 months (87.3% vs. 82.8%) and the whole period (85.3% vs. 83.7%). The proportion of days where antidiabetic drugs were used also was significantly higher (92.3% vs. 89.2%; 90.8% vs 86.6%). CONCLUSIONS: Providing free medications to diabetic patients raised their observance and may have improved patient outcomes.