As is consistent with national trends, the Montgomery County (Maryland) Public School System is exploring the use of instruments other than multiple-choice tests for high-stakes testing. This paper presents information on racial, ethnic, and gender differences in performance on the various types of tests being administered in the district. Sharing such information among school systems will help in the evaluation of new types of assessment. The six assessments used in the study were: (1) a mathematics multiple choice test given to grades 3 to 8; (2) a mathematics short answer test for grades 3 to 8; (3) a locally developed mathematics extended answer test for grades 4, 6, and 7; (4) a reading multiple choice test for grades 3 to 8; (5) a language arts extended answer test for grades 4, 6, and 7; and (6) the Maryland School Performance Assessment program for grades 3, 5, and 8. There were no meaningful differences in mathematics performance by racial and ethnic group across the different types of test studied. Nonmultiple-choice reading and language arts assessments favored nonwhite students. Nonmultiple-choice tests, whether in mathematics or language arts and reading, favored females over males. The largest difference between students on reduced-cost or free meals and others was in reading and language arts, where lower income students had substantially larger gains when moving from multiple-choice to nonmultiple-choice assessments. (Contains one reference and three tables.) (SLD)