The authors report on an exploratory pilot project implemented in a large city on the U.S.-Mexico border. The project focuses on a curricular design in a secondary school where beginning-level English as a second language students are taught English language arts in Spanish. The researchers sought to determine whether these students made greater gains on the Standard English Language Proficiency (SELP) exam than those students taught the same class primarily in English. Four classrooms participated in the study; the intervention class and three different control classes. A one-way analysis of variance was conducted with SELP gain scores as the dependent variable and the schools as the between-subjects factor. Follow-up "t"-tests and a multiple regression analysis suggest that less Spanish used in the classroom accounts for higher SELP gain scores. The authors discuss the challenges that were faced in implementing the project and caution that the study has several limitations. (Contains 3 tables and 3 footnotes.)