Sixteen Hispanic Spanish/English bilingual children (6 boys and 10 girls) participated in a single-subject design study. Their chronological ages ranged from 8 to 9.5 years. The classroom teacher identified all the children as "academic at risk" on the basis of a history of poor academic performance in spelling and low scores on the Metropolitan Achievement Tests (G. Prescott, I. Balow, T. Hogan, & R. Farr, 1978). The teachers assigned the students to each instructional condition according to a randomly selected sequence of instructional order. The 3 instructional interventions were (a) competitive team peer tutoring, (b) cooperative team peer tutoring, and (c) standard teacher-led instruction. The results of the study showed that although team competition and team cooperation resulted in higher levels of correct responding relative to the standard teacher-led condition, cooperative team peer tutoring resulted in the highest rate of correct responding. Practical implications of the findings are discussed. (Contains 2 figures.)