A column experiment was performed to investigate the hydraulic conductivity reduction with microbial growth in the riverbed during riverbank/bed filtration. Four columns were filled with glass beads, and varying flow rates were obtained in the columns by applying varying hydraulic heads. The microbial growths and the hydraulic conductivity reductions were measured at different column depths. Results show that the flow rates decreased exponentially along with the development of the biomass formed both on the surface and in the intermediate pore. The major factor for the flow rate reduction was the development of sludge cake of about 5 mm thickness on the riverbed surface rather than the intermediate clogging. The time needed was about 30days for the sludge cake development, and the thickness did not increase beyond 5 mm with further operation. The hydraulic conductivity of the cake was measured as cm/sec. The intermediate clogging was expected not to vary according to the changes in the operational conditions of the riverbank/bed filtration. It was also observed that the time needed was about 5 months for the intermediate clogging to develop completely and the biomass distribution along the column depth was independent of the applied head differences. The depth of the active microbial zone was about 50 cm from the surface. The model proposed by Ives and Pienvichtr (1965) was a good description for the relationship between the biomasses and the hydraulic conductivity reductions, of which the p value was 1.68. The microbial concentration of the intermediate biomass was calculated as 7000 mg-biomass/L-total volume.