A concurrent validation analysis was conducted for the recently developed Test of Spoken English (TSE), using as an external criterion the Foreign Service Institute (FSI) direct proficiency interviewing procedure. Use-related validation data for the TSE were obtained as a predictor of the communicative effectiveness in English of non-native English-speaking teaching assistants in U.S. colleges and universities. TSE and FSI tests were administered to 134 foreign teaching assistants at nine participating institutions. The TSE subscores were somewhat more reliable than those of the FSI, and exhibited a greater degree of discriminant validity. In the use-validation phase of the study, FSI and TSE scores of 60 non-native English-speaking teaching assistants were entered as predictor variables in multiple regression analyses, using as criterion variables student ratings of the instructor's spoken language. Both TSE and FSI scores were very effective predictors of student ratings of the instructor's speaking proficiency. Somewhat lower but properly directed weightings were found for the prediction of more global aspects of the teaching performance (e.g., overall effectiveness of the instructor). It is concluded that both the TSE and FSI can predict the probable communicative facility in spoken English of non-native teaching assistants in instructional settings. (Author/SW)