Socios En Salud uses directly observed therapy to treat a majority of the multidrug-resistant tuberculosis in Peru. The nurses play an important role in this community-based model as the patients' primary care givers. Since nurses, rather than physicians, are involved in patients' daily care, we developed a nurse-order entry system to test whether such a system would improve the accuracy and quality of medication data. We compared regimen information from patient electronic medical records, paper charts and pharmacy records. After a two-month training period on the new system, we conducted the trial for 52 days in two of Lima's six geographic treatment areas, and re-reviewed the three sources of medication data. We measured the error rates after the trial period and found there was no significant difference in the control group's (Lima Este), error rate (8.6% vs. 6.9%, P=O.66) after the trial. The intervention group (Lima Callao), however, showed a significant drop in the error rate (17.4% vs. 3.1%, P=0.0074) after the same time interval. Additionally, the nurse expressed satisfaction with the order entry system and its ease of use. The decrease in error rates and user satisfaction regarding the system are promising measures of our order entry system's success. © 2004 IMIA. All rights reserved.