Current theory and experimental research suggests that children's discovery of false beliefs at around 4 years of age allows the development of intentional deception. Anecdotal evidence of earlier lies has been dismissed with the argument that they may be ‘blind’ learned strategies rather than genuine deception. This paper presents two studies of everyday deception in comparison with false-belief task performance in young children. Study 1, a longitudinal study of 24 children, shows that the variety and incidence of everyday deceptions reported by mothers did not relate to success or failure on a battery of false-belief tasks, either between different children or over time in the same children. In Study 2 the deceptions of a 21/2-year-old child over a 6-month period were shown to be varied, flexible, context appropriate and too complex to be ‘blind’ learned strategies. It is argued that children's deceptive skills develop from pragmatic need and situational exigencies rather than from conceptual developments; they may learn to lie in the same way as they learn to speak.