The main objective was to describe parameters and physicochemical factors of activated carbon related to the adsorption capacity of three adsorbates: Phenol, benzoic acid, and salicylic acid. Two multivariate data analysis methods were used: Partial least square (PLS) and principal component regression (PCR). PLS showed better agreement between estimated and experimental values and using this method, equations were developed to predict the removal capacity of each adsorbate. The adsorption capacity of activated carbon in relation to benzoic acid, salicylic acid, and phenol was predicted with a standard error of validation of less than 6%. Surface acidity was the most important parameter affecting the adsorption of aromatic compounds by activated carbon.