This paper examines unintentional code-switching in Dutch learners of English as a second language and the relevance of these data for models of bilingual speech production. It focuses in particular on bilingual lexical access, bilingual lexicon storage of L1 and L2 words and phonological and morphological encoding processes.