The major problem for this research was to identify differences in personality characteristics among secondary teachers classified by performance levels, with the intent of improving selection procedures. To accomplish this task, data were collected from three personality inventories which were administered to 177 stratified, randomly-selected teachers who had been evaluated as either meeting or exceeding district standards by their supervisors. The teachers represented various teaching areas, ages, and levels of educational training;Responses from the 177 subjects on each of the three personality inventories were subjected to three two-way multivariate analyses of variance. After computing the Manova, the Wilks-Lambda was used as the multivariate significance test of group differences. If that test were significant, subsequent univariate F-tests were computed on the main effects and the interaction effects. Whenever the results were found to be significant, Tukey's Honestly Significant Different test was used to delineate group mean differences;The findings inicated there were statistical differences in three general areas: between teachers who were evaluated as exceeding district standards when compared to those who met; among teachers of various academic disciplines; and between teachers who were evaluated as exceeding district standards as compared to those who met district standards from various academic disciplines;The results of the study revealed that the 16 PF personality inventory shows the most promise in its use as a tool for the purpose of identifying significant personality characteristics of teacher applicants.