The objectives were to evaluate temporal variation in 137Cs in trout and charr from Crummock Water, Wastwater, Ennerdale Water, and to compare levels in trout from six lakes (above three + Windermere, Loweswater, Devoke Water) and charr from four lakes (above three + Windermere). Fish were caught between June 1986 and October 1988, chiefly with fyke nets and gill nets. Overall means for 137Cs in trout were markedly higher than those for charr in the three lakes. Values for individual trout exceeded 1000 Bq.kg-1 in Ennerdale and 500 Bq.kg-1 in Wastwater and Crummock Water, but values for charr never exceeded 350 Bq.kg-1. 137Cs content increased with fish weight for trout from Wastwater and therefore values were scaled to a standard weight in subsequent analyses. No similar relationships were found for trout in Ennerdale or Crummock Water, or for charr in all three lakes. Monthly mean values for 137Cs in trout followed similar temporal patterns in the three lakes and a parabola provided a simple model for these changes. Chernobyl deposition occurred early in May 1986 and maximum values in trout were in December 1986 in Wastwater, January/February 1987 in Ennerdale Water and March 1987 in Crummock Water. The exponential rate of decrease from these maxima did not differ significantly between the three lakes, and ecological half-lives of 137Cs in the trout differed slightly, but not significantly, between lakes : 180 days for Wastwater, 194 days for Crummock Water and 249 days for Ennerdale Water. The equivalent value for Ennerdale charr was lower at 132 days, and it was impossible to fit models or calculate half-lives for Wastwater and Crummock charr. Maximum monthly geometric mean 137Cs values in trout from six lakes and charr from four lakes were related to the initial concentration of 137Cs in both water and sediment, and the maximum monthly geometric mean 137Cs values obtained from the routine water samples. The relationship was curvilinear and approximated, but was not identical to, an asymptotic curve. No similar relationships were found with mean Ca or K levels in the lake water. Metabolic and diet differences may explain the lower 137Cs levels in charr compared with trout and differences for the same species from different lakes may be due to limnological variations and différences between lakes in the nature of their sediments and catchment soils. Both points are discussed in detail.