This study analyzes how two bilingual classroom language distribution approaches affect classroom language use patterns. The two strategies, separate instruction in the two languages vs. the new concurrent language usage approach (NCA) allowing use of both languages with strict guidelines for language alternation, are observed on videotapes of a Title VII demonstration project in San Antonio, Texas. The tapes were of students in primary grades and were made during one school year. Analysis of the classroom talk focuses on teachers' functional language use, measured by utterances coded according to function. The results provide some evidence that major concerns about the adverse effects of concurrent language use in bilingual classrooms are unfounded. No distortions in functional distribution patterns were found in the NCA classrooms, and the distributions of functional language were very similar for the two approaches. Fears of excessive switching and encouragement of intrasentential code-switching were not substantiated. NCA teachers did tend to favor English, but the overall ratio of Spanish to English use was not excessively unbalanced. Evidence was found that the fundamental assumption of the NCA, that students tend to follow the teacher's lead in language choice, is borne out. (MSE)