Saturated and unsaturated sand and soil column experiments were conducted to study the complex interaction between the effects of biological and hydrological factors on the transport of bacteria through a porous medium. These experiments were conducted with continuous input of bacteria and substrate at the inlet to reflect the groundwater contamination caused by leaking septic tanks and leach pits. Experiments were conducted with metabolically active and inactive Escherichia coli. Cell surface characteristics and batch experimental data for bacterial attachment were correlated with the transport behaviour in continuous column studies. Normalized breakthrough concentration for metabolically inactive cells (C/C0 = 0.74 in sand) was higher than that for active cells (C/C0 = 0.68 in sand) owing to change in cell surface characteristics. A similar trend was observed in the case of transport through soil columns. There was an increase of 29.5% in the peak C/C0 value at the outlet when the flow velocity was increased from 0.0535 cm/h (C/C0 = 0.61) to 0.214 cm/h (C/C0 = 0.79) in case of sand columns. However, this difference was only 20% in case of soil columns. Peaknormalized concentrations at the outlet were less in soil column as compared to those in sand column because of lesser grain size. Unlike the earlier studies with pulse input, present experiments with continuous input of metabolically active bacteria along with substrate indicated that the normalized concentration at the outlet increased with increased concentration at the inlet. It was found that unsaturated conditions led to more retention of bacteria in both sand and soil columns. In case of sand columns, the normalized concentration at the exitreduced to as much as 0.46. It was also found that the existing mathematical models based on macroscopic advection–dispersion–filtration equations could satisfactorily simulate the bacterial transport except in a case where the substrate was added to the bacteria in the column studies.