The performance of Degussa P-25 titanium dioxide photocatalyst in a water environment and two different forms of operation was studied. The finely divided powder was used in a suspended bed (slurry) reactor, and the same catalyst was physically supported on quartz sand and employed in a fluidized bed reactor. Both reacting systems were operated under conditions of a fully irradiated photoreaction space (FLP reactor) and the comparison was made for the same mass concentration of catalyst per unit reactor volume (250 ppm) using different combinations of bed expansion and surface coverage of the support. A tubular black light lamp was placed inside an annular reactor, where the photocatalytic oxidation of a model compound (oxalic acid) was investigated. Results were analyzed in terms of initial reaction rates per unit reactor volume and per unit of the liquid phase volume inside the reactor. Likewise, apparent quantum efficiencies were evaluated for the initial conditions, using the concept of apparent captured radiation power. It was found that the efficiency of the slurried photocatalyst is approximately five times larger than in the supported form. Among the different fluidized bed operations the highly expanded bed, with 100% coverage of the support surface, was found to be the most convenient condition for the immobilized titania.