ABSTRACT People daily perform tasks and talk about them. What is striking is that, although students perform many tasks in the classroom, they are rarely asked systematically to describe in detail how they proceed in performing them. The author proposes a self-observation and self-report procedure that will assist teachers in discovering their students' learning strategies and, as a by-product, may generate knowledge about the second language learning process. Transcriptions of segments of interviews with four students illustrate the questioning procedure and principles involved in eliciting students' strategies. Analyses of interview segments underscore a lack of congruence between what students are thought to be doing and what they actually are doing.