Bilingual compound verbs (BCVs) are documented in various languages and are common in codeswitching between English and South Asian languages. It has been suggested that BCVs have no monolingual equivalent, and are generated by a ‘third system’ independent of the two languages. BCVs have also been cited as evidence of language convergence, and as a strategy employed by dominant bilinguals to circumvent lexical gaps in one language. BCVs were common in narratives from four to six-year-old Panjabi-English children in Huddersfield, UK. BCVs are argued to be based on analogy with Panjabi monolingual compound verbs, and to be unrelated to language convergence or language dominance. Instead, BCV use relates to two types of codeswitching in the data: one utilising the simplest structures from both languages, the other drawing more fully on the two languages’ grammatical resources. It is suggested that BCVs enable children with limited overall bilingual competence to ‘do codeswitching’.