In three experiments, using the auditory moving window task, bilinguals operating in single (English) or dual (Spanish and English) language communities listened to successive sound segments of sentences presented one at a time. In Experiments 1a-b, sentences were in Spanish and the critical target was either a code-switch (|’pik(ə)ls|) or a borrowing, in which an English target was pronounced in Spanish (|pikos|). Experiments 2a-b compared code-switched versus non-switched targets within Spanish sentences. Experiment 3 used sentences in English and critical targets were in Spanish. Context (low/high constraint) and word frequency (low/high) were manipulated. Results for Experiments 1a-b revealed that code-switches took longer to process than borrowings. Taken together, findings from the three experiments suggested that code-switched language results in a processing cost in which the bilingual’s linguistic system demands more memory and time to successfully integrate the code-switched information into the sentence. Word frequency and context, as predicted by the featural restriction model, affected the processing of the code-switched targets.