In this study, ion-imprinted polymer was prepared for manganese(II) by formation of binary (1-(2-pyridylazo)-2-naphthol) complex in methanol (porogen) following copolymerization with 4-vinylpyridine (VP), ethyleneglycoldimethacrylate (EDMA), and 2,2'-azobisisobutyronitrile (AIBN) as a functional monomer, cross-linking agent, and initiator, respectively. The standardized effects of the independent variables and their interactions were also investigated using Pareto chart. Results of the two-level fractional factorial design (2(4-1)) based on an analysis of variance demonstrated that only pH of solution, amount of polymer, and adsorption time were seen to be statistically significant. For optimization of manganese removal from aqueous solution by manganese-imprinted polymer (Mn-IP) a three-factor, three level Box-Behnken design techniques combining with response surface methodology (RSM) was used on 15 different experimental data obtained in a batch study. The significance of independent variables and their interactions were tested by the analysis of variance (ANOVA) with 95% confidence limits. The optimal conditions were 9.7, 44.4 mg, and 19.1 min for pH of solution, amount of polymer, and adsorption time, respectively. Under the optimized experimental conditions, the detection limit of the proposed procedure followed by ICP-OES was found to be 0.6 mu g L-1. Furthermore, the distribution ratio and selectivity coefficients of manganese and other selected ions were investigated. The method was applied to the determination of manganese in water samples.