An experiment evaluated the impact of an integrated formal, analytic, and functional communicative approach to second language teaching in French immersion on French language proficiency. The impetus for the study arises from research indicating that immersion children show persistent weaknesses in their grammatical skills despite the fluent, functional proficiency they achieve in their second language. The experimental materials consist of a curriculum unit focusing on the conditional and involving students in planning an imaginary space colony. The materials were designed to teach the conditional by providing opportunities for students to use this form in natural, communicative situations; by reinforcing their learning with systematic linguistic games; and by encouraging metalinguistic awareness. Pre-, post-, and follow-up tests of oral and written French were administered to grade 7 early French immersion experimental and control classes (315 students from four school districts in the metropolitan Vancouver, British Columbia). Results showed that the experimental group performed significantly higher in writing in both post- and follow-up testing. Although these results were not found for speaking, examination of the individual class data revealed greater and more consistent growth in speaking for the experimental than for the control classes. (MSE)