The flow behaviour of impression materials during impression taking was first investigated, and reference made to the pressure which occurs in the compound. Individual trays of plastic and shellac-bases which could be softened by heating, and a cold-curing tray material were then also examined: (1) for their distortion by slight loads upon the tray walls; (2) for their ability to retain their shape when stored, and (3) their distortion when temporarily warmed to body temperature. It was found that no tray materials, although graded according to type and thickness, can be considered rigid under experimental conditions, since distortions which were in some cases considerable could be measured. In laboratory experiments, the behaviour of the tray was then measured while taking impressions of models and the precision of the impression was examined at the same time. This confirmed the suspicion that the precision of an impression is to a considerable extent related to the tray material used. The smallest alterations were found when using cold-curing plastic material of a sufficient thickness (greater than 2 mm).