Thermodynamics dictate that ideally the process of heat transfer obtaining within a feed heater should be reversible. Modern high-pressure feed heaters are, however, costly units and so it is also important that they should be economically efficient heat exchangers. These demands are in conflict and, in an effort to achieve highest thermal efficiency with lowest capital cost, the conditions for optimum design, the heat transfer and thermodynamic aspects must be a matter of compromise. It is essential, therefore, that the designer should have to hand means of readily determining the effects that changes in operating conditions, in processes within the system, will have on the efficiency and cost of the proposed plant. First, the thermodynamic aspects are dealt with; the ideals of regenerative feed-heating systems are briefly outlined. The ‘availability’ concept is suggested as a means of readily determining the effects that changes in operating conditions have on the efficiency of the plant as a whole. Generalized equations governing the heat-transfer processes obtaining within a desuperheating-type feed heater are next formulated. The large number of dependent variables operating in these equations, together with the range of possible boundary conditions, render any optimization process an exceedingly laborious task, involving as it must much repetitive computation. This is the type of problem most advantageous for solution on a digital computer and, without entering into the field of computer programming, an effort has been made in the paper to present the problem and its solution in generalized terms.