This paper has a double aim: i) to empirically establish the functions naturally expressed by any and those of one of its Spanish counterparts, specifically cualquier(a) so as to identify possible cross-linguistic transfer; ii) to illustrate a high-performing methodological procedure. A set of tools, among them an ad hoc tertium comparationis consisting of a set of cross-linguistic labels, a parallel corpus (P-ACTRES) and a reference corpus (CREA) are used to explore: i) the uses of any in context and the resulting translation environments served by cualquiera; ii) the degree of matching between translated cualquier(a) and its non-translated usage in standard European Spanish. The corpus-based procedure follows basically Krzeszowski's contrastive model (1990) with the addition of a 'target language fit' stage (Chesterman, 2004). The analysis shows different behaviour in translated and nontranslated Spanish: Cualquier(a) is underused as a translation option for 'existential' any and acquires a new function, 'negative', which is not a possibility in non-translated language for the same contexts. The analysis also corroborates the usefulness and the replicability of the methodological procedure.