Two contextualized degree of bilingualism measures, one designed to assess the extent to which each language is used, the other to assess relative proficiency in the two languages, were administered to 34 bilingual children of Puerto Rican background who attended a parochial school in Jersey City. The children reported that they used more Spanish, when talking to other bilingual Puerto Ricans, in the contexts of family and neighbourhood, than they did in those of education and religion. Their relative proficiency scores were in general agreement with their usage scores: the greatest difference between English and Spanish proficiency scores being observed for the domain of education and the smallest difference for the domain of family. In recent years there has been increasing recognition of the need to view bilingualism, not as a global capacity, but as one which could be described in terms of various components (3, 4). This view has led to the considera tion that bilingual proficiency might vary over a range of social settings. For example, a bilingual individual might be more proficient in one language when discussing matters of an academic nature and more pro ficient in another language when talking about household matters. Drawing upon this assumption, Cooper and Greenfield (1, 2 developed a series of instruments designed to measure degree of bilingualism in various domains or institutional contexts in which language behaviour occurs, e.g., family, education, religion. In the work reported in the present paper, two contextualized measures of degree of bilingualism were adapted for use with children. One measure was designed to tap bilingual