The purpose of this study was to develop an instrument to measure manual praxis by assessing areas of function considered important in learning and planning movement; use of tools, imitation of gesture, and motor sequencing. Participants included 362 healthy children aged between 3 and 12 years. Their parents completed a questionnaire on hand skills. Performance on all tasks improved with age. However, transitive gestures reached an early ceiling effect at around 5 years of age. Assessment of manual praxis from the parental questionnaire correlated with the child's directly measured competency during a test of object use (P<0.01) but not with the ability to imitate transitive gestures or enact a motor sequence with a 'novel' task. Children from schools where a higher number were eligible for free school meals because of low family income had more difficulty with imitation of gesture and motor sequencing to verbal and picture commands (P<0.001, P<0.05, P<0.05). They were neither disadvantaged in tool use nor in appreciation of extracorporeal space. The pattern of errors made in tool use and motor sequencing differed from those described in adults with acquired dyspraxia. These results suggest that assessment of manual praxis in clinical populations should include tool use, imitation of gesture, and motor sequencing and that they should not necessarily be regarded as measuring a unitary function.