Performance of calcium peroxide (CaO₂) in removing glucocorticoid (GC) contamination was investigated with triamcinolone acetonide (TA) as the representative target compound. An experimental design based on the response surface methodology (RSM) was applied to assess the individual and interaction effects of several operating parameters (CaO₂ dosage, initial pH value, reaction time, and TA concentration) on TAremoval. Based on the results of RSM, the optimal conditions were obtained as follows:CaO₂ dosage at 4g/L, initial pH=5.7, reaction time=15.7 min, and initial TA concentration=0.06 mmol/L. Moreover, additional experimental result obtained under optimal conditions (82.8%) was found to be very close to the predicted value (88.6%). These results indicated that the response model could be an effective method to describe the removal of GC by CaO₂. Finally, the role of reactive oxygen species (ROS) in GC degradation was tested by electron paramagnetic resonance spectroscopy (EPR). This study contributes significantly to the improvement and better understanding of CaO₂ oxidation and GC treatment processes.