The main objection against conventional net energy systems is that owing to variation in gain composition and the different energy contents of protein and fat, the efficiency of energy gain cannot be regarded as a growth constant. The present approach shows that the separate accommodation of protein and fat in predicting ruminant nutritional requirements can easily be achieved, since growth energy retention efficiency can be replaced by protein and fat synthesis efficiencies, together with an augmentation of maintenance with the cost of protein turnover. The synthesis efficiency of protein (kPS) is taken to be kPS = (qL/qM)(6/7), with 6/7 the synthesis efficiency of protein, qL the metabolizability of the diet at an arbitrary level (L) of intake and qM the metabolizability of the diet at maintenance. The correction (qL/qM) allows for the usual evaluation of ruminant diets at the maintenance level of intake. The synthesis efficiency of fat from fermentation of digestible fibre is kFF = 1.018qM or kFF = 1.287kg, without the necessity of adjustment by qL/qM, since evaluation of metabolizability at maintenance is incorporated in the relationship between kFF and qM and where kg denotes growth energy efficiency. Maintenance estimated from fasting heat production or intake at zero energy retention should be augmented by the cost of protein turnover from (PB/6) ÷ (qL/qM), with PB/6 = 102.7 kJ/kg (FW)0.75 per day for cattle and PB/6 = 78.1 kJ/kg (FW)0.75 per day for sheep, where PB denotes protein breakdown and FW fasted body mass. Alternatively, with knowledge of the degree of protein maturity, body protein turnover can be incorporated in a theoretically derived estimate of protein retention efficiency. The effective energy system can also be improved by employing theoretical protein retention and fat synthesis efficiencies or by equivalently replacing protein retention efficiency by protein synthesis efficiency in conjunction with the augmentation of maintenance heat production by the cost of protein turnover. A comparison between average growth energy efficiencies shows excellent agreement between estimates of the present theory and those of the UK Agricultural Research Council (ARC) and the US California Net Energy System (CNES), with degrees of maturity together with protein and fat gain ratios that seem typical of original experimental conditions. This implies that the present approach should do at least as well as the ARC or CNES, but can be expected to do better with reasonable accuracy in estimating the degree of protein maturity or maintenance augmentation and the composition of energy gain. The relationship between conventional growth energy efficiency and the synthesis efficiency of fat from digestible fibre allows the accumulated information of net energy systems to be transferable to the new methodology.