Semi-batch experiments were conducted to investigate the effects of catalyst type, pH, initial dye concentration and production rate of ozone on the catalytic ozonation of the dyes, namely Acid Red-151 (AR-151) and Remazol Brilliant Blue R (RBBR). The used catalysts were alumina, 25% (w/w) perfluorooctyl alumina (PFOA), 50% (w/w) PFOA and 100% (w/w) PFOA. The results showed that the overall percent dye removal after 30 min of the reaction was not affected significantly by the catalyst type. However, highest COD reduction was achieved by ozonation with alumina for AR-151, and 100% PFOA for RBBR at pH 13. The behavior of COD reduction with the increasing amount of perfluorooctanoic (PFO) acid amount can be explained by the enhancement of catalytic activity of PFOA with alkyl chains. For both of the dyes, the highest dye and COD removals were reached at pH 13. The overall dye reduction after 30 min of ozonation was almost independent of the initial dye concentration at relatively low values while at the higher concentrations, it changed with the initial dye concentration for both of the dyes. Similarly, COD reduction changed on a limited scale with the increasing initial dye concentration from 100 mg/L to 200 mg/L. however, an increase of initial dye concentration to 400 mg/L decreased the COD reduction significantly. All the studied production rates of ozone were sufficient to provide almost 100% dye removal in 30 min, whereas the COD removal percentage was increased gradually by the increasing ozone input to the reactor. The reaction kinetics for the ozonation of each dye with and without catalyst was investigated and discussed in the paper. (C) 2007 Elsevier B.V. All rights reserved.