Recent research on Content and Language Integrated Learning (CLIL) has confirmed that CLIL learners clearly outperform non-CLIL learners (Dalton-Puffer 2011). Although lexicon has often received pride of place in CLIL research, it is striking that studies have rarely been pushed beyond the word level. This paper therefore investigates the acquisition of Dutch posture verb constructions (PVs) in French-speaking (non-)CLIL learners. As in other Germanic languages, Dutch frequently uses compulsory PVs (staan [to stand], zitten [to sit] and liggen [to lie]), whereas French resorts to neutral verbs (être [to be]) (Lemmens 2002). Thanks to the content-based CLIL approach, which provides more opportunities to learn target vocabulary in meaningful situations, we expected to find qualitative and/or quantitative differences between the two learner groups. Data were collected from French-speaking (non-)CLIL secondary school sixth graders with Dutch as target language. They had to describe pictures displaying different types of environments (e.g. a bedroom, a clothing shop) featuring several objects located in varying manners/places. Our data show that CLIL learners use significantly more PVs than non-CLIL learners, but this does not mean that the use of PVs in CLIL learners can yet be regarded as ‘near-native’. Quite unexpectedly, CLIL learners still remain strongly influenced by their mother tongue: as compared to native Dutch, they display an overall underuse of PVs, concomitantly with an overuse of neutral verbs to locate objects. Although we notice some confusion between the different PVs in both groups, the CLIL learners use the appropriate PVs significantly more often than the non-CLIL. Other quantitative and qualitative results will be discussed during our talk. References Dalton Puffer, C. (2011), Content-and-Language Integrated Learning: From Practice to Principles? In: Annual Review of Applied Linguistics 31, 182–204. Lemmens, M. (2002). The semantic network of Dutch posture verbs. In: Typological studies ...