Eight feeding trials on cattle showed that the response in live-weight gain was curvilinear when protein or energy level was varied independently in rations based on poor quality roughages to which cereals and groundnut were variously added to give a range of protein and energy intakes. Live-weight gain was increased by 0-12 lb/day for each 0-1 lb additional digestible crude protein/day at levels of intake below 0-7 lb/day and by 0-03 lb/day at higher levels. It was increased by 0-4 lb/day for each additional 0-1 lb starch equivalent daily/100 lb live weight at total intakes below 0-9 lb daily/ 100 lb and by 0-2 lb/day at higher levels. Over the range 0-5-2-0 lb digestible crude protein daily and 0-7-1-2 lb starch equivalent daily/100 lb live weight the effects of these nutrients were additive.A Latin square nitrogen balance trial demonstrated that a portion of the nitrogen from a protein supplement was retained in the body despite the excretion of the greater part of it in the urine and a further part in the faeces. Nitrogen retention was increased by a supplement of readily available energy through a decrease in urinary nitrogen loss. Nitrogen balance was not increased by a supplement of a fibrous energy source because this induced an increase in faecal nitrogen loss.In a further feeding trial milk yield was found to be affected in a similar manner to growth. At a yield of 30 lb/day the response per lb starch equivalent added to the diet was 0-5 lb milk including 0-05 lb solids not fat and to an increase of 0-1 lb digestible crude protein daily it was 0-25 lb milk including 0-025 lb solids not fat. T~NTR O'DTTrTTO'N' similar result with pregnant heifers. These observations suggest that energy intake orientates the In earlier experiments (Bailey & Broster, 1957; response curve to protein relative to absolute rate of Broster, Tuck & Balch, 1963) we examined the live-weight gain. variation in the rate of live-weight gain of heifers Few of the published experiments on levels of given different amounts of protein in rations calcu-feeding for dairy cows have dealt with the question lated to be isocaloric. The response was curvilinear: of independent variation of protein and energy the increase in rate of gain from additional protein intakes. They have been concerned largely with the decreased as the basal level of intake rose.