This paper discusses experiments undertaken in 1969 and 1970 using natural beach material introduced onto Chesil Beach from other areas. The 1969 trials used approximately 17,000 quartz granulites; 1970 trials used two batches of approximately 6000 quartzite-jasper conglomerates and one batch of over 5000 basalts. In 1969 one site was used; in 1970 two were chosen to provide comparisons. Trial searches showed variability in recovery rates reflecting both the nature of the introduced material and the effect of the size of the background pebbles. The experiments showed: i) Rates of movement for the quartz granulites at Wyke reached 343m per day but after 165 days the farthest travelled was only 3952m from its origin. This reflects both the proportion of time that material is out of circulation by being above or below the zone of wave action and transport opposite to the prevailing direction from time to time. The 1970 experiments showed different periods of burial for the two trial sites and less consistent direction of movement at Portland where wave approach is more nearly at rightangles to the shore. Data show that lateral movement of individual pebbles is not necessarily greater under storm conditions. ii) Recovery patterns indicate concentrations of experimental material at high and low water marks. There are also concentrations along the beach which may be related either to anomalies in beach curvature or to the time available for longshore movement. iii) Maximum recovery rates occur under long, low ground swell conditions. Under these circumstances, where the pebbles are larger than the natural background material they are 'rejected' from it and lie on the surface. As a consequence they are available for 'attack' by beach processes for a disproportionate length of time. Recovered material is not necessarily typical of that injected initially being differentially selected depending upon the particular wave conditions. The percentage of high specific gravity basalts recovered differs from that of other types of added material since the basalt pebbles work their way into the beach only to be exhumed at a later date. iv) There is a relationship between pebble size and longshore movement. Correlations have been found in both directions. Short diameter is most sensitive to changes in wave period and direction. v) Shape sorting and shape abrasion do not appear to play an important part on this beach. This is possibly the effect of the pebble traction carpet and the high wave energy at the site. vi) Size sorting is probably affected in the vertical plane by phase difference . Longshore sorting is most likely achieved by a vector imparted by the direction of wave approach. This vector is more effective at Wyke than at Portland, since it is more oblique to the coast and the traction carpet is composed of finer material.