Deception is a controversial aspect of theory of mind, and researchers disagree about whether it entails an understanding of the false beliefs of one’s opponent. The present study asks whether children with delayed language and delayed explicit false belief reasoning can succeed on explicit deception tasks. Participants were 45 orally taught deaf children with varying language delays aged 4.5–8 years and 45 hearing children aged 3.5–6 years. Participants received a battery of language, executive function, deception, and both verbal and low-verbal false belief tasks. The result reveal a dissociation of deception and false belief tasks: the deaf children are on par with their hearing peers on deception games, but show significant delays in false belief tasks even when the language demands are made minimal. Furthermore, different skills are predictors of success for the two types of task in the deaf children: language, and in particular complement syntax, is the best predictor of false belief reasoning; but executive function skills, especially inhibitory control, are the best predictors of deception. It is argued that deception at this level can be handled by behaviour rules without reference to mental states.