Accumulation of pore-filling materials produced by bacteria, such as insoluble extracellular polymeric substances (EPS) can modify the properties of soil matrix and decrease the porosity and hydraulic conductivity, is referred to bioclogging. Numerous laboratory experiments and field-scale applications have shown the substantial reduction in hydraulic conductivity of porous media by the growth of biofilm producing bacteria. However, the effect of grain size in biofilm production and hydraulic conductivity have not yet been investigated. Therefore, this study explores the effect of grain size in hydraulic conductivity with a model bacterium Leuconostoc mesenteroides. Two kinds of sands having different grain size were selected and used as a host sand: (1) fine sand (Ottawa F110) and (2) coarse sand (Ottawa 20/30). An acrylic (polycarbonate) column equipped with two pressure transducers and a differential pressure transducer was used to measure hydraulic conductivity of soil specimen. Two experiments with model bacteria Leuconostoc mesenteroides was conducted in two separate phases. During the first phase (~25 days), the hydraulic conductivity reduced by more than one order of magnitude regardless of the grain size. After the hydraulic conductivity reached a plateau, the neutral pH growth medium containing 0.1 M potassium phosphate buffer was injected into the columns (Second phase). The hydraulic conductivity began to decrease again in both columns. Thereafter, the feasibility of bioclogging for field application was investigated by observing the hydraulic conductivity variation with bacteria originated from embankment. The reduction of hydraulic conductivity caused by natural soil bacteria was also observed during the experiment which was carried out for ~47 days. The hydraulic conductivity at the middle part of soil column decreased by approximately two orders of magnitude which is assumed to be the contribution of inhomogeneous biofilm formation. After the hydraulic conductivity reached plateau, the acrylic (polycarbonate) column was dismantled to measure the quantity of biofilm formation. From the measurement, it was observed that biofilm was formed more around inlet part of column. Furthermore, the biofilm pore saturation and thickness were estimated with hydraulic conductivity variation by using several analytical models. While pore saturation showed great gap with different model, biofilm thickness showed relatively similar range. The pore saturation estimated with mass measurement was much smaller than that estimated with analytical models.