Carrots were grown annually from 1954 to 1962 in a rotation of vegetable crops in a long-term factorial manurial experiment. The experimental treatments were applied to the same plots before each crop so that possible cumulative effects of the treatments could be measured. Farmyard manure (FYM), at 20 tons per acre per crop, gave consistently higher yields than fertilizers alone and the increases were greater in dry years than in wet. The difference in yield between the manured plots and the fertilizer-treated plots did not increase consistently after 1959, thus indicating that a saturation level of the factors causing the yield increases had been reached. In the presence of FYM, applications of N, P and K fertilizers had only small effects on carrot yields. In the absence of FYM, more than 31 lb. per acre N as Nitrochalk depressed yields, whereas both phosphate and potash fertilizers increased them. Plants from the FYM plots contained about twice as much potassium and about 30% more phosphorus than those from the fertilizer-only plots. These differences are discussed in relation to the increased yields on the FYM plots and to the effects of the fertilizers alone on the yield and composition of the plants.