ABSTRACT: Oil palm trunk (OPT) represents one of the five main oil palm biomass wastes with high carbon content that can be economically converted to a large surface area, porous activated carbon (AC) adsorbent to treat palm oil mill effluent wastewater in Indonesia and Malaysia. In the first portion of this work, the design of the experiment was used to determine the optimum set of synthesis parameters required to maximize the iodine number of AC [i.e., Brunauer-Emmett-Teller (BET) specific surface area indicator] prepared from OPT via chemical activation route using H3PO4. The iodine numbers of AC and AC yield were probed as the impregnation ratio, the activation time, and the activation temperature were varied in the range of 0.28-3.47, 5.68-69.32 min, and 379-521 °C, respectively. An impregnation ratio of 2.29, an activation time of 6 min, and an activation temperature of 450 °C were identified as the optimum set of synthesis parameters. In the second portion of the work, the AC synthesized using the optimum parameters were then characterized and tested as an adsorbent for tannin. N2 sorption results revealed that the AC exhibits type IV isotherm, that is, contains micropores and mesopores and displays a relatively high BET specific surface area of 1657 m2 g-1. Adsorption equilibria isotherms for tannin adsorption onto the AC were collected at three different pH of 2, 4, and 6 and were nonlinearly fitted using Langmuir and Freundlich isotherm models, where the Langmuir isotherm gave better fitting than Freundlich. The higher adsorption capacity at lower pH can be explained in terms of the absence of electrostatic repulsion interaction between the AC surface and the tannic acid species as suggested by the point of zero charges (pHpzc) of 4.8 and an increasing ionization of tannic acid with pH rise between 4 and 7. Adsorption kinetics data were also obtained at four different pH of 2, 4, 6, and 8 where the nonlinear pseudo-first-order model best fitted the kinetic at pH of 2 and the nonlinear pseudo-second-order model represented the kinetic best at the remaining higher pH, which suggests that tannin adsorption onto AC occurred by physisorption at pH of 2 and by chemisorption at pH of 4, 6, and 8.