The present work aims to contribute to our understanding of the underlying dimensions of language learning strategies in foreign language contexts. The study analyzes alternative factor structures underlying a recently developed instrument (Tragant and Victori, 2012) and it includes the age factor in the examination of its construct validity. The target population consists of middle- and upper-grade learners of English distributed in two samples (n[subscript 1] = 550 and n[subscript 2] = 1425). Exploratory factor analysis and item analysis were initially conducted to be followed by confirmatory factor analyses and multiple-groups factor analysis. The instrument is a 55-item questionnaire based on a 6-point Likert-type scale measuring students' reported frequency of strategy use. Results support a correlated two-factor structure with a shortened scale of 17 items reflecting "skills-based deep processing strategies" and "language study strategies," offering empirical evidence for the distinction between deep and surface clusters of strategies. Multiple-groups factor analysis showed that this model held for both middle- and upper-grade students, and upper-grade students were more likely to use the more advanced skills-based deep processing strategies and less inclined to use language study strategies than middle-grade students. The brevity of the scale and parsimonious factor structure enhance the questionnaire's utility for research and classroom evaluation. (Contains 2 tables and 1 figure.)