The purpose of this study was to devise quantitative methods for the assessment of praxic skills of the upper limbs by developing a computerised task which permits each component of a sequence of actions to be timed precisely. Furthermore, two versions of such a quantitative measure were developed to investigate the relationship between meaningful and meaningless movements. The praxic skills of 35 patients with Alzheimer's disease (AD) and 75 healthy controls were assessed on two 3-item sequential movement tasks involving either meaningful or meaningless actions. A qualitative rating scale assessment of gesture imitation and pantomime was also administered. AD patients were significantly slower than controls on both the sequential movement tasks. Indeed, the correlation between AD patients' abilities on the novel and traditional tasks provided evidence that the sequential movement tasks constitute valid measures of praxis. Within the AD population, disease severity was also found to have a minimal and inconsistent influence upon praxis. The apraxia assessment results are considered in relation to the debate over whether apraxia constitutes an early or late feature of AD, and also to theoretical claims about the cognitive neuropsychological deficit underlying ideational apraxia.