A controversial issue in bilingual research is whether in the early stages of L2 learning, access to the conceptual system involves mediation of L1 lexical representations [Kroll, J. F., & Stewart, E. (1994). Category interference in translation and picture naming: Evidence for asymmetric connections between bilingual memory representations. "Journal of Memory and Language, 33", 149-174] or a direct route from the L2 word [Altarriba, J., & Mathis, K. M. (1997). Conceptual and lexical development in second language acquisition. "Journal of Memory and Language, 36", 550-568; Finkbeiner, M., & Nicol, J. (2003). Semantic category effects in second language word learning. "Applied Psycholinguistics, 24", 369-383]. The main goal of this paper is to study, in a child population, whether the creation of conceptual representations for L2 words is possible, even after only one session of learning of the L2 vocabulary. Furthermore, we do so by examining the efficacy of two different L2 learning methods: L2-L1 association learning vs. L2-picture association learning. A translation recognition task was employed to test whether there was a difference between a semantically related pair and an unrelated pair across conditions (i.e., a "semantic interference" effect). Results showed a significant semantic interference effect--a conceptual effect--in children after just one vocabulary learning session. Importantly, the L2-picture method produced a greater semantic interference effect than the L2-L1 method. The implications of these findings for models of bilingual memory are examined. (Contains 2 tables and 1 figure.)