Fish were collected from the intake screens of the Oldbury Power Station in the Severn Estuary in each week between early July 1972 and late June 1977 and at least twice monthly between early January 1996 and late June 1999. The annual catches, after adjustment to a common sampling effort, demonstrate that the abundance of fish at Oldbury was far greater in the 1990s than 1970s, mainly due to marked increases in the numbers of certain marine species, such as sand goby, whiting, bass, thin-lipped grey mullet, herring, sprat and Norway pout. These increases may reflect the great improvement that occurred in the water quality of the Severn Estuary between these decades. The only species that declined markedly in abundance was poor cod. Modest declines in flounder and River lamprey paralleled those occurring elsewhere in the UK. The species composition in the two decades also differed, reflecting changes not only in the relative abundances of the various marine estuarine-opportunistic species, which dominated the ichthyofauna, but also in those of the suite of less abundant species in the estuary. The cyclical changes undergone each year by the species composition of the fish fauna of the Severn Estuary reflect sequential intra-annual changes in the relative abundances of species representing each of the marine, diadromous and freshwater categories. New approaches have been developed to test whether or not large sets of correlations between patterns of recruitment amongst abundant marine species (internal correlations), and between those patterns and salinity and water temperature within the estuary (cross-correlations), were significant. The correlation profile analyses found no evidence that the annual recruitment strengths of these species were either intercorrelated, or correlated with either one or a combination of both of the above environmental variables. Yet, the timings of the recruitment of these species into the estuary were intercorrelated, i.e. a slightly earlier or later than normal immigration by one species in a given year was paralleled by the same trend in other species. However, this association in recruitment times could be linked neither to salinity nor water temperature within the estuary, nor to a combination of these two variables. These results indicate that, while the factors that influence the annual recruitment strengths of the juveniles of different marine species vary, inter-annual differences in the phasing of events that regulate spawning times and/or larval dispersal influence, in the same direction, the times when marine species are recruited into the estuary.