Elderly people have a higher risk of developing diseases than young people because of physiological changes which occur with aging. Therefore, elderly often use many medications, which makes the incidence of medication errors higher. Also, polypharmacy is the leading cause of drug interactions. The aim of the research was to monitor the incidence and type of potential clinically significant drug interactions in 179 elderly patients (over 65 years) hospitalised at the Clinical Department of Internal Medicine, University Hospital Dubrava, and compare those interactions found in best possible history of medications used prior to admission with the ones in the list of medications prescribed at hospital discharge. The research also determines the most common number of potential clinically significant interactions as well as the drugs that most frequently participate in those interactions and the most common drug-drug interactions. Lexicomp drug interaction software was used for the screening of potential interactions. The data was collected by interviewing the patients, their relatives, family physicians and pharmacists as well as by studying their medical documentation. Total number of potential interactions found in the discharge medication lists was bigger than in the list of drugs prior to admission (1218 vs. 1094), but there were less potential interactions of D and X level of clinical significance. On the contrary, the number of patiens in whose medication list at least one potential clinically significant interaction of the level D and/or the same of the level X was bigger in the discharge list than in the list of medication prior to admission (118 vs. 108). The results show the high need of screening and medication reconciliation for the elderly hospitalised patients, which implies an involvement of clinical pharmacists in the team of healthcare professionals.