The study examined the relationship between learner characteristics and achievement in French as a second language in a sample of 391 adult immigrants completing a 900‐hour course of classroom instruction in Montreal. A multivariate analysis of variance was used to assess the relative contribution of nine learner variables to individual differences in performance on an FSL achievement test. Teachers' evaluations of subjects as “good learners” or “poor learners” served as the dependent variable in a multivariate ANOVA comparing the two groups. Results of the various analyses indicated that subjects who have more schooling, a higher degree of nonverbal reasoning ability, make greater use of French outside the classroom, are more competent in English, and display a greater degree of field independence are more likely to benefit from formal language instruction. Higher levels of illiteracy and classroom anxiety, coupled with greater age, were related to learning difficulties. The paper underlines the importance of multivariate data analysis techniques and proposes the use of graphical data analysis techniques for researchers exploring multivariate data sets with small groups of