Incident reporting systems are ubiquitous, but are limited by low clinician reporting rates and the fact that they identify a relatively small percentage of adverse events. In this feasibility study, a handheld computer-based system for voluntary reporting of medication errors was deployed among physicians and nurses at four academic hospitals. The system demonstrated promise as a supplement to traditional error reporting systems, although even with the increased ease of use, only 16% of clinicians in the study actually reported an error during the study period.