The photodegradation of three commercially available dyestuffs (C.I. Reactive Black 5, C.I. RB5, C.I. Direct Yellow 12, C.I. DY12, and C.I. Direct Red 28, C.I. DR28) by UV, UV/H2O2 and UV/H2O2/Fe(II) processes was investigated in a laboratory-scale batch photoreactor equipped with an 16W immersed-type low-pressure mercury vapour lamp. The experimental results were assessed in terms of absorbance and total organic carbon (TOC) reduction. The initial concentration was kept constant at 100 mg 1(-1) for all dyes. Initial results showed that,, color removal efficiencies by UV or H2O2 alone were negligible for all dyes. Almost complete disappearance of C.I. RB5 (99%) and DY 12 (98 %) in UV/H2O2 process was possible to achieve after 60 min of irradiation. The maximum color removal efficiency of C.I. DR28 after 60 min of irradiation, however, was only 40% and reached a maximum value of 70% after 120 min of irradiation. Corresponding mineralization efficiencies were 50, 55 and 7-12%, respectively. The addition of Fe(II) to the system, so-called the photo-Fenton process, greatly enhanced the color removal, the efficiencies being 98, 88 and 85% for C.I. RB5, C.I. DY12 and C.I. DR28 only after 5 min of irradiation. Corresponding mineralization efficiencies were 98% for 45 min irradiation, 100% for 60 min irradiation and 98% for 90 min irradiation, respectively. However, marginal benefit was less significant in the higher range of both H2O2 and Fe(II). Furthermore, decreases in both decolorization and mineralization were observed at higher concentrations of oxidant and catalyst due to the scavenging effect of excess H2O2 and OH* radicals. The degradation of all dyes was found to follow first-order reaction kinetics. (C) 2004 Elsevier B.V. All rights reserved.