An approach to the optimization of a solar energy conversion system which involves treating the system as a series of subsystems, each having a single cost determining variable, is proposed. Optimization techniques can be used to determine designs for each subsystem for constant values of the cost determining variable. Subsequently, the allocation of a financial resource amongst subsystems to achieve an optimal performance can be determined. The application to an ammonia-based thermochemical system with direct work output is discussed and possible subsystems are identified. The subsystem consisting of the exothermic reactor has been studied in detail. For this subsystem, the ratio of available catalyst volume to thermal power level is held constant whilst the exergetic efficiency is maximized. Results are presented from a determination of optimized reaction paths using dynamic programming techniques.