The goal of the present study was to explore the pattern of lexical connections between translation equivalents represented in the cognitive system of an Arabic-Hebrew bilingual. To achieve this goal, repetition priming effects (reaction times (RTs) and accuracy measures) were compared between translation equivalents in the two forms of Arabic, Spoken Arabic (SA) and Literary Arabic (LA), with those between translation equivalents in Hebrew and LA, in native Arabic students living in Israel. While these students master Hebrew as well as they master LA, they consider it as a second language. LA targets were preceded by cognate and non-cognate translation equivalents in SA or Hebrew either immediately (Lag 0) or separated by four unrelated words (Lag 4), in a lexical decision task. The participants were requested to make a word/non-word decision regardless of language. Across languages, morpho-phonemic similarities augmented the repetition priming effects. At Lag 0 this effect was larger when the primes were in Hebrew than when they were in SA. At Lag 4, however, the priming effect between Hebrew and LA was significantly reduced (and even absent for non-cognates). In contrast, the priming effect between Spoken and Literary Arabic was only slightly attenuated. The conclusion was that, despite the intensive daily use and psychological proximity, Spoken and Literary Arabic are represented as first and second languages in the cognitive system of the native Arabic reader, entertaining distinct lexica. However, the strength of the lexical associations between translation equivalents is influenced by the frequency of concomitant use.