In some hospitals, clinical pharmacists review the medication to find drug-related problems (DRPs) in acutely admitted patients. We aimed to identify the nature of identified DRPs and investigate factors of potential importance for the clinical implementation of pharmacist suggestions. In 100 randomly selected medication review (MR) notes, we retrospectively evaluated the clinical implementation and classified (1) timing and communication of the review; (2) DRPs and related suggestions for the physician; and (3) DRPs' potential clinical relevance to patients as 'beneficial', 'somewhat beneficial', 'no relevance' or 'other relevance'. Of 327 DRPs (0-13 DRPs per patient), 42% were implemented. The clinical implementation was higher if the MR note was made prior to (instead of after) the physician's admission, and even higher if the suggestions were communicated verbally (instead of only in writing) to the physicians (44% versus 79%, p < 0.05). The clinical relevance of the DRPs was either 'beneficial' (16%), 'somewhat beneficial' (43%), 'no relevance' (22%) or 'other relevance' (19%). The 'beneficial' DRPs had a higher clinical implementation (53%) than 'no relevance' (34%) (p < 0.05). The most frequently implemented suggestions were based on DRPs concerning 'indication for drug treatment not noticed', 'inappropriate drug form' and 'drug dose too low', with implementation rates of 83%, 67% and 63%, respectively. In our sample, the pharmacist's MR suggestions were only implemented by physicians in 42% of the cases, but review prior to physician contact and verbal communication of the suggestions, higher clinical relevance and specific types of DRPs were associated with a higher implementation rate.