A study at Brigham Young University (Utah) investigated the feasibility of computer-assisted language placement testing in higher education. Benefits and problems of this approach for test administration, individualization of item selection, and recordkeeping were examined. Four steps were followed in production of a test for Spanish placement: development of an item bank; determination of a suitable psychometric methodology; selection of a test delivery system; and validation and implementation. A 1,443-item bank was created, reduced to 1,100 items, and incorporated into 5 test forms. Tests were administered to 199 students from lower-division Spanish courses, then submitted to conventional statistical and Rasch analyses. Items were indexed to 51 difficulty levels. After final review by faculty, the 1,050 items remaining were coded, and 9 items at each difficulty level were included in the final version of the placement test. The test was then designed so the computer would present to each examinee items in the estimated range of his/her ability and inform students immediately of their probable class placement. The test is delivered on a standard IBM or compatible personal computer with two disk drives. Student records are stored and can be printed. Initial test implementation was under way at the time of the report's writing and expansion of such testing was anticipated. (MSE)