This paper presents a novel, knowledge-driven approach for pointing device prescription that aims to enhance the prescription process via OSCAR, a Clinical Decision Support System. OSCAR was constructed in a three-stage process: developing a pointing-device ontology; formulating IF-THEN matching rules, and constructing a web-based user interface. Evaluating OSCAR's usability through the System Usability Scale reflected a high satisfaction rate with a mean score = 80.9 (SD = 10.6). OSCAR's effectiveness was evaluated through the prescription results which showed no significant difference between experts using the traditional approach and novices using OSCAR. These results indicate that OSCAR is able to support novice clinicians in the AT prescription process to a level that is comparable to experts. The main contribution of this research is in the development and evaluation of a CDSS which bridges the gap between existing general models for pointing device prescription and a process that can support expert and novice clinicians. Although the system is currently designed for adaptation of a specific device (physically controllable pointing device) it can be expanded to include other AT devices. The results will also serve as an example to demonstrate the utility of using a CDSS in other clinical areas.