The putative bilingual executive advantage has been argued to stem from lifelong experience with executively demanding language behaviors, such as switching between the two languages. However, studies testing for possible associations between language switching frequency and EF in bilinguals have yielded inconsistent results. One reason for this could lie in the methods used that have evaluated the frequency and type of language switches with retrospective self-reports, as well as in problems in reliability and convergent validity of the executive tasks. By using Ecological Momentary Assessment (EMA) as a reference point for self-reports of language switches, we examined the validity of general retrospective self-reports of language switching. Additionally, we examined associations between language switching and EF using multilevel models. Our results indicated that the commonly used retrospective self-reports of language switching may lack convergent validity. However, we found tentative evidence that contextual language switches, assessed with EMA, may be associated with better inhibitory control, set shifting, and working memory.