Good music aptitude tests can help discover talented students and improve music education by finding the weak and strong points of students. Especially Korean students need a music aptitude test that can measure both Western and Korean music aptitude. In this study, we have developed a music aptitude test, called Korean Music Aptitude Test (KMAT), for the 5th to 9th grade Korean students. KMAT was developed from August 1996 to September 1998, through fourteen times of pilot studies. The subject of this study consists of the 4th, 5th and 6th grade students from 36 elementary schools, and the 7th, 8th and 9th grade students from 14 junior high schools. The total number of the students is 7,297. KMAT was administered to 2,514 students from 14 schools in August and September 1998. KMAT consists of rhythm and tonal sub-tests, In the rhythm sub-test, students are required to answer same or different after listening to pairs of rhythm patterns without pitch. In the tonal sub-test, students are required to answer same or different after listening to pairs of 5-note tonal patterns without rhythm. The rhythm test items use 3/4, 4/4, 6/8, 12/8 meters. The tonal test items are developed by considering the direction and shape of melodies. The rhythm test reliability was .85. The tonal test reliability was .77. The inter-correlation between the rhythm sub-test and tonal sub-test is .41 for the 5th grade and .20 for the 6th grade. It means that the rhythm test and the tonal test for Korean students have are not closely related, and so both tests are needed. We obtained the concurrent validities of KMAT for the 5th and 6th grades. Concurrent validity is the correlation between the KMAT scores and the scores of a music achievement test. The music achievement test for the 5th grade was developed by an elementary school teacher for the 5th grade and the author. The achievement test for the 6th grade was developed the same way. The concurrent validity of KMAT for the 5th grade is .55 and that for the 6th grade is .40. Both correlations are statistically significant(p<.01). But to obtain better validity, music achievement tests whose validities are demonstrated are needed. Further required studies include finding correlations with other standardized music aptitude tests, and obtaining predictive validity of KMAT.