Do bilinguals use the same logical representations in processing both languages they know if the languages they know diverge on the preferred interpretation of a scopally ambiguous sentence? We investigated this question by studying effects of cross-linguistic influence and priming in the comprehension of scopally ambiguous *all...not* sentences. In this investigation, we focused on Estonian-English bilinguals, because these sentences are interpreted differently in English compared to Estonian. Across four sentence-picture matching experiments, we firstly observed that bilingual logical representations can be primed within languages (both from L1-to-L1 and from L2-to-L2) and between languages (both from L1-to-L2 and from L2-to-L1). This finding indicates that bilinguals make use of shared logical representations in processing the languages they know, and that logical representations do not specify language-specific biases in the assignment of scope. Secondly, we observed that bilinguals experience bidirectional cross-linguistic influence in the construction of logical representations. This finding suggests that bilingual representations of implicit knowledge about scopal preferences is integrated between the languages they know.