Three studies were conducted, one in autumn with decreasing water temperatures (11 degrees to 5 degrees C), one in spring with rising temperatures (7 degrees to 12 degrees C) and one in summer in relatively stable temperatures (mean 13.2 degrees C). Three diets (T1, T2, T3) containing 470/100, 450/150, 430/200 g kg(-1) of protein and fat, respectively, were fed to provide equivalent energy intakes. Actual mean feeding levels were: Trial 1, 13, 12, 11 g kg(-1) body weight; Trials 2 and 3, 16, 15, 14 g kg(-1) body weight, respectively, for T1, T2, T3. In Trial 2 a fourth treatment (T4) was diet T3 given at 16 g kg(-1) and in Trial 3 a fourth treatment was diet T4 formulated to contain 410/230 g kg(-1) of protein and fat and fed at 13 g kg(-1) body weight. Initial weights of trout were 170 g in Trial 1, 145 g in Trial 2 and 215 g in Trial 3. Initial pond loads and stocking densities were 184, 40, 30 kg and 12.4, 18.4, 13.8 kg m(-3). All three trials were of 6 weeks duration. It is concluded that in the spring and summer trials protein intake was adequate on the diet with the highest energy:protein ratio but that protein may have been limiting at the lower feeding levels which occurred with falling temperatures in the autumn trial. Observed improvements in feed conversion ratios with the higher energy diets were in line with expected values in Trial 2 and were better than expected in Trial 3. The effects of high fat diets on fillet and offal composition were also examined. It was found that fillet and offal fat levels on a fresh and dry matter basis were significantly increased as increasing amounts of fat were fed. (C) 1997 Elsevier Science B.V.