This report describes the preliminary clinical trial and evaluation of ProtolSIS, a prototype case-based reasoning (CBR) system for medical decision support. ProtolSIS learned the domain of ultrasonography and body computed tomography from 200 consecutive cases of actual requests for imaging procedures. ProtolSIS was tested by presenting it with 100 new imaging-procedure requests. ProtolSIS correctly classified 72% of the imaging-procedure requests overall, and 84% of the cases in the last two test series. We evaluated ProtolSIS in terms of performance, utility, and acceptibility to physicians. CBR can be applied successfully to the selection of diagnostic imaging procedures and holds potential for use in clinical decision support aids. Further work is necessary to realize a clinically useful system.