Medical errors are common in hospitals, and research is always needed to find ways of reducing these. This study attempts to address three gaps in this field. First, the factors leading to the reduction of mental workload and its relationship with the reduction of prescribing errors by improving electronic prescribing (e-prescribing) usability have not been empirically examined before. Second, the past research in the field of e-prescribing usability lacks robust theoretical models. Third, there are no existing studies to examine the direct influences of user interface consistency and error prevention with the reduction of mental workload and prescribing errors. A quantitative survey method was used to collect data from 188 community physicians. The partial least squares path modeling technique was applied to analyze the data. Prescribing errors were reduced by improving the information quality, user interface consistency, system ease of use, and mental workload reduction. Mental workload is reduced by ease of use, error prevention, and consistency. No significant relationships between prescribing error reduction with error prevention and also between information quality with mental workload reduction were found. The designers of e-prescribing should improve the error prevention and consistency of the system and make it easy to use if they wish for the system to reduce users' mental workload. They should also improve the system information quality, ease of use, and consistency if they claim that their system reduces physicians' prescribing errors. The system should also reduce users' mental workload to meet this objective.