The photocurrent has been regarded as the most efficient method to evaluate the photocatalytic activity of composite photocatalyst. A relationship is commonly recognized: the higher the photocurrent is, the higher the e(-)-h(+) separation efficiency is, and thus the higher the photocatalytic activity is. In this paper, we use the TiO2/WO3 composite as a model heterojunction system to argue about this viewpoint, and try to explore a more comprehensive relationship. Our research reveals that there is no direct relation between photocurrent, charge separation efficiency and photocatalytic activity. It is shown that the photocurrent combined with the carrier mobility is essential in the analysis of the charge separation efficiency. In addition, the charge separation efficiency cannot be responsible for the photocatalytic activity, since it does not include the information on the redox ability of the charge carrier. It is confirmed that the photocatalytic activity which is a pluralistic concept includes three levels of degradation efficiency, initial reaction rate and mineralization ability. The difference in each level is determined by the band structures, the matched band potentials and the target pollutants. Therefore, we also propose that the introduction of pollutant ambience is needed in photocurrent test. The ratio of photocurrent in air to that in pollutant ambience can be used as an index to evaluate the initial reaction rate of photocatalytic activity. (C) 2012 Elsevier B.V. All rights reserved.