Sequence-alignment methods which have been recently introduced in time-use research analyse the cross-sectional as well as the sequential information embedded in space - time activity-diary data. As these methods have been developed to trace the evolutionary history of random mutations of the elements in biological sequences, they are not appropriate to measure the distances by which activities change their sequential order in activity diaries. In this paper we report on the development of a position-sensitive sequence-alignment method for activity analysis. First, the problem is specified, and the key concepts to analyse the problem are addressed. Then, the principles of the new method are introduced. The method is illustrated with data on activity patterns collected in the Netherlands. The paper concludes with a critical discussion of the basic assumptions of the method and avenues of future research are identified.