he influence of the halogen upon the TiO2 photocatalytic degradation of p-halophenols in water has been investigated. Phenol was used as the reference compound. Compared with its value for phenol, the apparent first-order rate constant of removal, k, was slightly but significantly higher for p-fluorophenol and p-chlorophenol, and slightly but significantly lower for p-bromophenol. For p-iodophenol, k was about half that of phenol. The relative values confirm that k is roughly correlated to the Hammett constant; this constant reflects the electron density on the aromatic ring and, accordingly, the reactivity towards electrophilic species generated by UV-irradiation of TiO2. All compounds were found to be poorly adsorbed on TiO2. Accordingly, k was not related to the differences observed in the very low adsorbed amounts. The detected aromatic intermediate products included hydroquinone (HQ), benzoquinone (BQ) and various halodihydroxybenzenes. HQ, BQ, 4-chloro-1,2 (and 1,3)-dihydroxybenzenes and 4-bromo-1,3-dihydroxybenzene were quantified. Mechanisms are tentatively suggested to interpret the differences in the degradation pathways of the p-halophenols. The organic intermediate products accounted for only a few percents of the total carbon during the degradation. The toxicity (1/EC50) measured by the Microtox test almost did not vary in the course of the degradation of phenol, p-chlorophenol and p-bromophenol until complete removal of these compounds. By contrast, the value of 1/EC50 was multiplied by ca. 2.5 when ca. 45% of p-iodophenol had been removed; concentrations of BQ higher than with the other p-halophenols are tentatively suggested to be at the origin of this increase. Interpretation of a surprising substantial increase in the 1/EC50 value when the removal of p-fluorophenol increased from 80 to 95% requires further investigation.