The suitability of diatom analysis for provenancing British late neolithic and early Bronze Age ceramics is examined. The Milfield Basin, Northumberland, was chosen for a test program, as an area with well documented clay sources, defined geology, and available pottery samples. The bulk of the pottery sampled was of local origin, or at least need not have been made elsewhere; one sherd, the only possible Beaker sherd sampled, seems to be an import. It is concluded that it is possible to use diatoms in this way, and some of the problems of the technique are discussed, together with possible solutions. Sampling and counting strategies are also discussed. -- AATA