The Common Achievement Test (CAT) in Japan, which will be implemented in 2005, involves a medical interview that is the core task to be completed by students during an Objective Structured Clinical Examination (OSCE). Standardized/Simulated Patient instructors (SPs), posing as patients in medical interviews, are trained in standard fashion in terms of expression of symptoms as well as the emotional affect of actual patients. Institution of appropriate training programs for SP instructors in the CAT is also necessary. We trained seven individuals to function as standardized patients (in-school SPs) during a three-day SP training program described in this article. Following completion of the OSCE, we conducted a comparison study among evaluations completed by the evaluators and two types of SP instructors. We observed high correlation, according to Spearman significance testing, between scores of evaluators and those of both newly trained in-school SPs and veteran SPs who had more than five years of experience. Correlation coefficients between the veteran SPs (r=0.77) and the in-school SPs (r=0.73) were nearly identical. These results suggest that our training program for SP instructors is an effective protocol, particularly with respect to reliability and efficiency.