A new computerized testing system, which facilitates the use of short-answer-type testing, has been developed. In this system, the question of a multiple-choice problem is presented first, and the options appear briefly on the request of the test taker. The crux of this manipulation is to force students to solve the problem as if they were solving an open-ended question but to respond in a multiple-choice format. To examine the system's usefulness as a learning aid, 2 experiments were performed. In Experiment 1, 140 6th-grade students were divided into 2 groups: The experimental group took the intervening computerized test in the computerized modified multiple-choice testing (CMMT) format, and the control group in the traditional multiple-choice format. In a paper postrecall test, the mean score was higher for the experimental group than for the control group. In Experiment 2, the result of Experiment 1 was replicated in a within-subject design. Possible mechanisms of the enhanced retention in the CMMT system are discussed.