We are investigating whether adult Spanish-English bilinguals engage shared or separate cortical regions to process simple arithmetic across their languages. Increasing evidence from the bilingualism literature suggests that bilinguals have language networks that are interconnected and highly interactive. However, some argue that math information is encoded and processed in the brain in a language-specific manner. Simple arithmetic facts, like multiplication tables, are thought to be language-specific memories because bilinguals typically learn and prefer to perform simple arithmetic in only one language. In support of this idea, bilinguals have been reported to be faster and more accurate in the language in which they learned math (LA+) than their other language (LA-), confirming a language bias for arithmetic. Previous studies on arithmetic processing have suggested that the intraparietal sulcus (IPS) is engaged during mental calculation in adults and children. The left superior and middle temporal gyri (STG/MTG) have been associated with representing memorized arithmetic facts, along with the inferior frontal gyrus (IFG) in retrieval of these facts. Critically, STG/MTG and IFG are also implicated in language comprehension. The STG/MTG become more active, and the IPS less active, as reliance on arithmetic facts increases and the reliance on calculation decreases. In addition, the IFG is more engaged in effortful retrieval when facts do not have a robust memory representation in temporal cortex, such as with less practiced large multiplication problems. Research to date has implied that engagement of these areas may be specific to LA+. Based on the bilingualism literature, we predict that both languages should engage overlapping temporal regions, reflecting interconnected language systems, even for arithmetic. Additionally, we predict that LA- might require additional processing and would therefore recruit additional brain areas, including IFG reflecting more effortful verbal retrieval or IPS for reliance on back-up calculation processes. This language difference would be predicted to be more pronounced for more difficult multiplication problems in comparison to easier problems.