ABSTRACT: Although the design of the study was exemplary, the number of classes and teachers was barely satisfactory, and “strategy” effects may have been confounded with text variables. The “traditional” strategy tended to be a “traditional-modified” one and may not have been as distinct from the “functional skills” strategies as might be desired. Technical flaws in the statistical analysis included failure to take account of repeated measures and questionable use of criterion-type variables as covariates, but the main results regarding “strategy” and language laboratory system are valid. Sometimes the analysis was not deep enough, as where there was a failure to rule out variation in student characteristics in connection with correlations between class means on achievement tests and teacher FL proficiency. The study is interpreted as showing that functional skills strategies and materials fail to put enough emphasis on linguistic content. The results on correlations between teacher FL proficiency and class achievement are not persuasive enough to justify abandonment of FL proficiency tests in teacher selection and certification.