The four purposes of this study were: (1) To compare two versions of a tailored testing model similar to one suggested by Kalisch (1974); (2) To identify levels of the variables within the two versions, which produce an efficient tailored testing procedures; (3) To compare, within each version, the results obtained when employing relatively small data bases with those obtained using large data bases; and (4) To determine, within each version, whether any of the tailored testing procedures are more efficient than obtaining the responses to all the items from each examineee. The models specified procedures by which the correctness of an examinee's responses to items of a test are predicted by requiring an examinee to answer as few test items as possible. The study employed computer simulations, using existing data collected from examinees who had answered all the test items. It was concluded that the model, and the levels of variables within the models, produce differential efficiency, dependent on the evaluation matrix. The testing procedures employing small data bases are not necessarily as efficient as the procedures employing large data bases, but for the evaluation matrices used in the study, the tailored testing procedures were more efficient than obtaining an examinee's response to every item. (RC)