This article proposes that a complex issue such as bilingualism gives rise to a need for complex research. Complexity theories, both in the psycholinguistic and educational fields, may inspire new empirical studies on bilingualism that will likely provide data otherwise unattainable through classic pre-test/post-test methods. The article also warns against hardcore educational egalitarianism which may seriously harm Content and Language Integrated Learning programmes if they become one-size-fits-all models. The article has been written in response to a reply to these authors' original article Lorenzo et al. (2010).