Bilingual memory representation or more precisely the way bilinguals restore words in their mental lexicon and retrieve them has intrigued psycholinguists in the past decade. Research on the psycholinguistic elements of bilingualism has increased at a 'dizzying rate' during the last years (Kroll & De Groot, 2005). The present study explores how bilinguals share semantic features of their L1 and L2 using a masked semantic paradigm. Target-prime pairs addressed in the study were cross-language translation pairs in two different directions from L1 to L2 and vice versa including abstract and concrete words. The results of experiment 1 showed a significant priming effect from L1 to L2 for concrete words. The results of experiment 2 showed a significant priming effect from L2 to L1) for abstract words. The results suggest that idea that L1 and L2 are represented using a similar lexico-semantic architecture in which L2 words are also able to activate semantic information. This is consistent with models assuming quantitative rather than qualitative differences between L1 and L2 representations.