An innovative classroom-based approach to reading instruction in the context of Spanish instruction was proposed. The effects of this instruction on the pronunciation ability of students were analyzed. The subjects were 30 adult missionary trainees who had no previous exposure to Spanish. The dependent variable was measured using two instruments. A posttest-only control group design was selected, using analysis of covariance, with Modern Language Aptitude Test scores as the covariate. Analysis of both dependent measures showed a significant treatment effect as well as a significant aptitude-treatment interaction. Results confirm Wilkinson's (1975) research regarding the improved effect on pronunciation ability of missionaries due to early reading instruction. Analysis of interaction showed reading instruction to be most beneficial to low-aptitude subjects, with high-aptitude subjects scoring more favorably overall. These findings suggest the need for the consideration of the presence of such an interaction in research evaluation. (Author/AM)