Reactive dye has generally been used in woven cotton fabric dyeing industries. Some treatments of several reactive dyes may produce more toxicity than the original dyes. The objectives of this study were to find the optimal condition on dye degradation efficiency of commercial reactive red dye 36 (DR36) and reactive violet dye 30 (DV30) using Fenton-like reaction, and to determine acute toxicity by static bioassay method under the optimal condition. The experiment was designed by Box Behnken Design (BBD), in which an initial pH, catalyst dosage and initial concentration of H2O2 were considered as independent variables. The results showed that only an initial pH solution was the principal parameter which influenced decolorization of the reactive dyes. Other factors were much less significant. The optimal conditions were found to be given by pH 3, 1 g/L of catalyst dosage, 27.63 mM of concentration of H2O2 for DR36, and pH 3, 1.35 g/L of catalyst dosage, 45 mM of concentration of H2O2 for DV30. Ninety percent of both decolorization were achieved in 30 min. Acute toxicity tests of the treated solutions using freshwater fairy shrimps (Streptocephalus sirindhornae) revealed that the shrimps survived longer than 24 h, indicating that the treated solutions were not acutely toxic. The average leaked iron, ADMI value and total organic carbon were found to be less than 10 ppm, 5 ADMI and 9.17 ppm respectively, in the treated samples. This research demonstrated an efficient method for decolorization of the reactive dyes with low acute toxicity.