OBJECTIVE:: To evaluate the effect of a Web-based calculator and decision-support system on infusion ordering errors and to estimate error frequency in pharmacy infusion preparation. DESIGN:: Data on ordering error frequency and typology were collected before and after implementation of an online infusion ordering system. Data on pharmacy preparation errors of infusions were collected. SETTING:: A children's hospital at an academic medical center. PATIENTS:: None. Data were abstracted from infusion orders. INTERVENTIONS:: Introduction of a voluntary-use Web-based calculator into infusion ordering workflow. Observation only. MAIN OUTCOME MEASURES:: Number and type of errors in handwritten and calculator-generated orders. Number and type of errors in pharmacy infusion preparation. RESULTS:: Before calculator deployment, 129 sequential handwritten infusion orders were collected over 5 weeks. After deployment, of 162 sequential infusion orders, 88% (142) were calculator-generated. Calculator-generated infusion orders contained 83% fewer (p < .001) orders containing one or more errors than handwritten orders. Calculator-generated orders contained no high-risk errors (incorrect decimal, dose, or unit of measure) when compared with handwritten orders and were associated with fewer pharmacy interventions. In 118 sequential pharmacy infusion preparations over 4 wks, there were no errors observed. CONCLUSION:: A Web-based calculator reduced significantly the total number of errors and eliminated all high-risk errors in the prescribing process for continuous pediatric infusions. With no observed errors in pharmacy preparation, this study provides data to support the use of computerized ordering as an independent safe and viable method for ordering continuous pediatric infusions.