This report consists of three sections, the first of which is a discussion of placement testing. It is maintained that the strengths and weaknesses of a tested population cannot be determined by a single composite test--yet, many institutions place students in beginning courses on the basis of such limited information. To assess basic capabilities assumed to be acquired in the lower level course, a test must be devised after the course objectives are known. National standardized tests are not always sufficient. By employing multiple correlations or multiple regressions (which are explained in more detail in section two of the report), the proper placement of a student can be enhanced. Data can be computer processed for the selection of the best predictors in rank order. It is suggested that regression equations developed from data collected in one testing session be applied to the data collected during the testing one calendar year later. Problems such as the contamination of input data (e.g., different grading standards used by various teachers) are discussed with suggestions being given for developing more accurate regression equations. Section three describes English placement at El Camino College (Calif.). Test scores from the Purdue Test, the Cooperative English Test, and SCAT were submitted to a computer which ranked the test scores in order of their effectiveness as predictors. Recommendations for the El Camino program are given. (DG)