This study has demonstrated the use of empirical modeling in resolving the effects of individual carbon properties on acid blue dye adsorption. Acid blue dye adsorption tests were conducted on activated carbons prepared from bagasse by physical (CO2) and chemical (ZnCl2, MgCl2 and CaCl2) techniques. Empirical models based on the carbon textural (surface area and pore size) properties and the surface chemistry inferred from heteroatom (C,H, N, and S) concentration and carbon surface pH were used to resolve the effects of individual carbon properties on acid blue dye adsorption. This form of analysis was conducted to optimize carbon preparation properties, forming the foundation for tailor-making adsorbents from bagasse suitable for acid dye adsorption. A series of statistical analyses (partial F-tests to establish the parameter significance) measured variants including the mean square error, r(2) and adjusted r(2), normality, and randomness of residuals, and formed the basis for testing the adequacy of these models. The empirical models suggest that a combination of suitable pore structure and distinct basic surface chemistry generated by sulfur- and nitrogen-based groups, which were also elucidated by Fourier transform infrared spectroscopy, is necessary to promote acid dye adsorption.