This paper examines the hypothesis that lexical item selection shows the degree of conceptual fluency in the target language. The case example used to support this assumption gathered production data in an immersion school in Barcelona, where the students' first, second and third languages are respectively Catalan, Spanish, and English.