At admission, unintentional medication discrepancies (UMDs) can occur and may led to severe adverse events. Some of them are preventable through medication reconciliation (MR). As MR is a time-consuming activity, a better identification of high-risk patients of UMDs is mandatory. The objective was to identify risk factors associated with UMDs at admission in an internal medicine department. This prospective observational study was conducted from April 2017 to June 2019. At admission, inpatients had MR to obtain a complete list of home medications. This list was compared to prescriptions made at admission. All discrepancies were classified as intentional or UMDs. Univariate and multivariate analyses to identify the risk factors associated with UMDs were performed. MR was performed on 1157 patients (70.1 ± 16.8 years old); 550 MR (47.5%) contained at least one UMD. More than half of the UMDs (n = 892, 65.6%) corresponded to drug omission. The univariate analysis showed that age (> 60 years old), “living at home”, medication preparation not performed by patient, medication-intake difficulties, number of sources consulted, MR duration, presence of a high-risk drug and the number of home medications were associated with UMDs. In the multivariate analysis, adjusted on the number of sources consulted, independent risk factors were “living at home” and the number of home medications. At admission to an internal medicine department, UMDs were frequent and associated with “living at home” and poly-medication. Our findings might help physicians to identify high-risk patients of UMDs since their admission.