Objective: to compare the prescription of medications to elderly users of the SUS with users of a Supplementary Health Plan in the context of the Beers criteria. Method: a quantitative documental analysis using two databases, obtained from surveys conducted in the same municipal region, was performed. In the first database, the sample included 532 people aged 60 years or older, of both genders, who were users of the SUS. In the second, the sample was 239 people aged 60 years or older, irrespective of gender, who were users of the principal supplementary health plan. Statistical analyses to compare the data of the two databases were performed using Pearson’s Chi-squared Test and the extension of Fisher’s Exact test, with a significance level α equal to 5%. Results: The results showed significant differences in the use of medications and polypharmacy among the users of the two service providers. Moreover, there were statistically significant differences between the service providers, with SUS users using more non-steroidal anti-inflammatory drugs (p=0.01), long-lasting sulfonylureas (p=0.02) and Nifedipine (p=0.01), and the users of the Supplementary Health Care plan using more musculoskeletal relaxants (p=0.01), estrogen (p=0.01), amiodarone (p=0.01) and Doxazosin (p=0.01), which are potentially inappropriate for the elderly. Conclusion: there are differences between having health insurance or not in terms of the profile of drug use, including in medications which are potentially inappropriate for use among the elderly. The use of information technology that centralizes the data of the elderly, both in the SUS and in Supplementary Health, could reduce inappropriate or unnecessary prescriptions.