The existing open ponding system for palm oil mill effluent (POME) treatment typically fails to eliminate phenolic compounds. The aim of this research was to develop an improved photocatalytic process that utilizes visible light to remediate phenol laden treated palm oil mill effluent (TPOME). The visible-light active core-shell structured black TiO2 (CS B-TiO2) was synthesized by a modified precipitation technique using glycerol as a renewable and environmentally friendly template. The photocatalytic process was first optimized using central composite design (CCD) in response surface methodology (RSM) for removal of model phenol and chemical oxygen demand (COD), and then used to remove the two substances from TPOME. Under non-optimized conditions, the synthesized CS B-TiO2 removed 76.05 of phenol, which is four times higher than the commercial TiO2 (P25). Through regression analysis, statistically significant quadratic polynomial models for phenol and COD removal efficiency were obtained. High R2 (0.9842) low mean square error (MSE, 0.086), and root mean square error (RMSE, 0.294) values indicated that the accuracy and generalization of both the optimization models were adequate. 90.73 and 94.43 of model phenol and COD were removed under optimal conditions of 1.27 g/L CS B-TiO2 loading, 0.06 mol/L H2O2 dosage, 30 ppm initial concentration of phenol, pH 7.2, and 180 min visible light irradiation. The removal efficiency of phenols and COD from TPOME matrix remained at 70.20 and 78.65 , respectively. This study confirms that an optimized photocatalytic process based on CS B-TiO2, which is sensitive to visible light, is an effective solution for the remediation of agro-industrial wastewater. © 2021 Elsevier Ltd