This study is aimed at comparing the effects of teaching English to Thai undergraduate teacher-students through cross-curricular thematic instruction program based on multiple intelligence theory and through conventional instruction. Two experimental groups, which utilized Randomized True Control Group-Pretest-posttest Time Series Design and statistical data analysis using MANOVA with repeated measures, t-test for independent samples, basic statistical analysis and figures, showed significantly higher development in English performance (reading, listening-speaking and writing), expressed significantly positive opinions at a much higher level on the learning program, and identified more experiences in using each and every aspect of multiple intelligences than what the two control groups did. The results thus show that prospective teachers whose English language performance was enhanced with integrated content areas can benefit their students for learning English across the curriculum.