Adverse drug events can occur as a result of handoffs in patient care. To reduce the possibility of this occurring, the process of medication reconciliation (whereby the patient's medication history is compared to current and previous medications to ensure accuracy) is becoming recognized as becoming increasingly important. To address this, computerized medication reconciliation tools have been developed. This paper describes a combined approach to evaluating the impact of such a tool. The approach has included both an artificial laboratory-based evaluation component (involving observing subjects interacting with standardized patient cases), as well as a naturalistic condition (involving real patient cases). The results indicate that there are differences in the way that subjects interact with the medication reconciliation tool, with significant differences identified in the amount of time spent and accuracy of medication documentation between physician and pharmacist users.