This study investigates the rate kinetics for BD and PGMEA oxidation by UV/O3 process. The reactor constructs of a 100 cm x 20 cm x 85 cm (L x W x H) stainless steel chamber, in which four vertical steel plates (20 cm x 65 cm, W x H) were inserted to establish a plug flow path for the flowing gas. The reactor has a total effective volume of 170 L. Each of the five compartments of the reactor is equipped with an individual UV irradiation system with a 3.0-cm x 15-cm (ID x L) quartz sheath that housed an UV lamp, and two electric UV power inputs of 0.147 or 0.294 W/L were obtained. The gas flows perpendicularly to the UV lamps in the reactor. The influent tested VOC concentration was adjusted to about 50 ppm, and the gas flows were controlled at the individual flow rate of 60 and 120 L/min. The effects of moisture content (relative humidity, RH), ozone dosage (initial molar ratio of ozone to the tested VOC, m) and UV volumetric electric power input on the removal of the tested VOCs are investigated in the study. Also, kinetic models of the tested VOCs by photolysis, ozonation and UV/O3 have been developed and confirmed with reference to the experimental data. According to the kinetic models, both photolysis rate and oxidation rate by UV/O3 are following the first order behavior with respect to the tested VOC concentrations which are low. The result reveals the absorbance for the reactions is weak absorbance under UV irradiation. The reaction rates are proportional to the UV electric power inputs in UV-initiated reactions. And the parameter, £i, which represents the ratio of OH radical consumption rate by the tested VOC to the total OH radical consumption rate, can be obtained by simulating the performance of experimental data of OH reactions. The experimental results reveal that for BD oxidation with a gas space time of 85 sec and RH = 40 ¡V 99%, BD photolysis did not occur at wavelength of 185 nm with UV electric power inputs of 0.147 and 0.294 W/L. The ozonation efficiency of BD reached 90% at m = 3.5, and RH had no influence on the removal efficiency of BD. The removal efficiencies by UV/O3 process reached 90% with m = 2.2 and 1.6 for UV power inputs of 0.147 and 0.294 W/L, respectively. The addition of ozone apparently encouraged BD removal efficiency by UV/O3 process. And the enhancement of ozone dosage (m = 0.5 ¡V 4.4) would promote the decomposition of BD more effectively than the enhancements of UV power input (from 0.147 to 0.294 W L-1) and RH (from 40 to 99%). For PGMEA photolysis in a batch reactor with volume of 1.188 L, the photolysis occurred at wavelength of 185 nm under UV irradiation. And the photolysis rate follows the first order behavior with respect to the concentration of PGMEA. But PGMEA photolysis did not occurred at UV wavelength of 254 nm. PGMEA ozonation was performed in the same batch reactor; and the removal efficiency of only 50% at m = 3.96 would take 35 min. So, PGMEA ozonation in the plug flow reactor did not be observed at the conditions of the gas space time of 85 sec and RH = 15 ¡V 99%. Besides, the photolysis of PGMEA was carried out at the above conditions. The removal efficiency of PGMEA by UV/O3 could reach 90% at the conditions of the gas space time of 170 sec, UV volumetric electric power input of 0.294 W/L and m = 2.9. And the enhancement of UV power input (from 0.147 to 0.294 W L-1) would promote the decomposition of PGMEA more effectively than the addition of ozone dosage (m = 1.05 ¡V 15.63) and RH = 15 ¡V 99%.