After preliminary surveys during 1977-78 more or less regular samples were collected from the River Thames about 500 m south of the town centre at Kingston-upon-Thames, at approximately weekly intervals, throughout the growing seasons 1979-1981.The taxonomic part of the investigation began with the separation of the centric diatoms into eighteen "types", (N.B. this term has no nomenclatural or taxonomic implication in this context), the cleaned frustules of which were visibly different when examined by light microscopy. An attempt has been made to correlate the observations obtained by light microscopy, stereoscan and transmission electron microscopy. Illustrations were prepared using the three methods to enable detailed comparisons to be made with descriptions in the literature. Some of the "types" described can be referred to species of Cyclotella e.g. C. compta (Ehr.) Kutz. and C. meneghin-iana Kutz. Others, especially those referable to the genus Stephanodiscus, are much more debatable. The ecological aspects of the investigation involved a determination of the numbers of centric diatoms present at the various times and an attempt to correlate these data with such factors as weather conditions and nutrient availability. Data were obtained from the MeteorologicalOffice, the Thames Water Authority and the North Surrey Water Co. After cleaning, samples were examined by light microscopy to determine their composition in terms of the percentage of different "types". Estimates were made of the population size of each of these "types" of diatoms in the various samples, so that seasonal changes in populations could be assessed. Different seasonal behaviour of some of the visibly recognisable "types" of Stephanodiscus suggests that these "types" do represent forms which, whether or not they represent different taxa, are at least physiologically distinct.