The present study examines two methods of using the Vineland Adaptive Behaviour Scales as a measure of behavioural change in people with Down syndrome who are at risk of developing Alzheimer’s disease. The first method uses the Vineland scales as the basis of a semi-structured interview and notes all areas of behavioural change identified by staff; the second method scores the Vineland scales using the basal rule outlined in the manual. The comparison of these two methods illustrated that using the second method highlighted a significant decline in scores for the group meeting the criteria for ‘probable Alzheimer’s disease’ on a number of domains between baseline and 12–24 months. However, this scoring method also appeared to miss more subtle changes in behaviour, which may be important early indicators of Alzheimer’s disease, which were picked up by the first method. The implications of the study are discussed.