Five experiments were carried out to gain insight into foreign language learning as it takes place in the classroom environment, and to create optimal learning conditions for beginning Swedish-speaking learners of English as a Foreign Language. The experiments examined receptive and productive comprehension of English vocabulary in Swedish-speaking students between the ages of 10 and 12. The students either had no formal English language instruction or had had less than one year of English instruction in school. The first experiment tested the uninstructed students' ability to identify aurally common English words and expressions, and the second experiment investigated the English-instructed students' ability to understand passages from fairy tales. The third experiment tested the uninstructed students' ability to make lexical inferences from a text in which Swedish-English cognates were maximized and "false friends" excluded. Experiment four required the English-instructed students to write as many words as possible beginning with a particular letter, with the language unspecified. The fifth experiment studied the effect on vocabulary learning of playing a computer game with English vocabulary, in which new vocabulary was defined on-screen. Analysis of the five studies' results suggests that Swedish-speaking students have an English vocabulary of varying size when they begin English language study in school, and that the quality of the vocabulary depends on the type of input available to each learner, with input on topics of interest learned more readily. Implications for instructional materials are considered. (MSE)