Process conditions for the direct solar decomposition of sulfur trioxide have been investigated and optimized by using a receiver-reactor in a solar furnace. This decomposition reaction is a key step to couple concentrated solar radiation or solar high-temperature heat into promising sulfur-based thermochemical cycles for solar production of hydrogen from water. After proof-of-principle a modified design of the reactor was applied. A separated chamber for the evaporation of the sulfuric acid, which is the precursor of sulfur trioxide in the mentioned thermochemical cycles, a higher mass flow of reactants, an independent control and optimization of the decomposition reactor were possible. Higher mass flows of the reactants improve the reactor efficiency because energy losses are almost independent of the mass flow due to the predominant contribution of re-radiation losses. The influence of absorber temperature, mass flow, reactant initial concentration, acid concentration, and residence time on sulfur trioxide conversion and reactor efficiency has been investigated systematically. The experimental investigation was accompanied by energy balancing of the reactor for typical operational points. The absorber temperature turned out to be the most important parameter with respect to both conversion and efficiency. When the reactor was applied for solar sulfur trioxide decomposition only, reactor efficiencies of up to 40% were achieved at average absorber temperature well below 1000 degrees C. High conversions almost up to the maximum achievable conversion determined by thermodynamic equilibrium were achieved. As the re-radiation of the absorber is the main contribution to energy losses of the reactor, a cavity design is predicted to be the preferable way to further raise the efficiency. Copyright (C) 2009 John Wiley & Sons, Ltd.