Bilinguals purportedly outperform monolinguals in non-verbal tasks of cognitive control (the ‘bilingual advantage’). The most common explanation is that managing two languages during language production constantly draws upon, and thus strengthens, domain-general inhibitory mechanisms (Green 1998 Biling. Lang. Cogn.1, 67–81. (doi:10.1017/S1366728998000133)). However, this theory cannot explain why a bilingual advantage has been found in preverbal infants (Kovacs & Mehler 2009 Proc. Natl Acad. Sci. USA106, 6556–6560. (doi:10.1073/pnas.0811323106)). An alternative explanation is needed. We propose that exposure to more varied, less predictable (language) environments drive infants to sample more by placing less weight on consolidating familiar information in order to orient sooner to (and explore) new stimuli. To confirm the bilingual advantage in infants and test our proposal, we administered four gaze-contingent eye-tracking tasks to seven- to nine-month-old infants who were being raised in either bilingual (n = 51) or monolingual (n = 51) homes. We could not replicate the finding by Kovacs and Mehler that bilingual but not monolingual infants inhibit learned behaviour (experiment 1). However, we found that infants exposed to bilingual environments do indeed explore more than those exposed to monolingual environments, by potentially disengaging attention faster from one stimulus in order to shift attention to another (experiment 3) and by switching attention more frequently between stimuli (experiment 4). These data suggest that experience-driven adaptations may indeed result in infants exposed to bilingual environments switching attention more frequently than infants exposed to a monolingual environment.