ObjectiveThe aim of this study is to determine the magnitude of medication discrepancies and its associated factors at transitions in care of a Specialized University Hospital in eastern Ethiopia.ResultsThis study enrolled 411 patients having at least one prescription medication. For each of the patient enrolled, a medication reconciliation process was accomplished between medication use history before transition and medication orders at the transition. A total of 1027 medications were reconciled and 298 of them showed discrepancies. From such medication discrepancies, 96 (32.2%) of them were unintended discrepancies. Patients admitted to surgical ward (adjusted odds ratio {AOR} 0.27 [95% confidence interval 0.10–0.74]) and on malnutrition therapy (AOR 0.13 [0.03–0.52]) had reduced likelihoods of medication discrepancies. However, patients on cardiovascular drug therapy (AOR 5.69 [2.4–13.62]) and who were hospitalized for more than 5 days (AOR 5.69 [2.97–10.9] {5–10 days}) had significantly increased likelihoods of discrepancies. Accordingly, one-third of the medication discrepancies identified were unintentional and these discrepancies were more likely to occur with cardiovascular drugs, in medical or pediatric wards and patients hospitalized for prolonged time. Therefore, this pharmacist-led medication reconciliation indicates the potential of pharmacists in reducing drug-related adverse health outcomes that arise from medication discrepancy.Electronic supplementary materialThe online version of this article (10.1186/s13104-018-3668-z) contains supplementary material, which is available to authorized users.