For predicting the adsorption performance of filters, the adsorption rate parameters, including the mass transfer capacity coefficient and the equilibrium concentration are required. Estimating these parameters involve a series of experiments on concentrations and amounts of the adsorbent, which is expensive and requires sufficient time for tests. This study aims to estimate the adsorption rate parameters more efficiently using computational fluid dynamics (CFD); moreover, it predicts the adsorption performance for differently shaped filters, made of the same material, using the predicted parameters. A filtration test is performed for a single column (the granular activated carbon packed bed), following which, the adsorption rate is calculated by fitting the CFD results for experimental results, particularly the mass transfer capacity coefficient and the adsorption constant, which determines the equilibrium concentration. The proposed CFD modeling method enables the adsorption rate prediction for activated carbon and the prediction of filter adsorption performance even in solutions with concentrations as low as 100 μg/L or less. This indicates that the proposed method aids in the adsorption rate and adsorption performance predictions for differently shaped filters with a simple filtration test (a water flow test) using a single column.