The HYDRUS-1D mobile–immobile water model (MIM) was used to evaluate the transport of fecal coliforms, Salmonella bacteriophage, and Br in 10 soils. At a flux of 5 mm h−1, a pulse of dairy shed effluent was applied to 30 large undisturbed lysimeters, followed by water irrigation. Soil types included clayey gley soil, clay loam, silt loam, silt loam over gravels, fine sandy loam, dune sand soil, pumice soil, and allophanic soil. Except for dune sand, modeling results showed lower mobile water contents and dispersivities for microbes than for Br, indicating the exclusion of microbes from smaller pores. The MIM-derived removal rates were in the order: volcanic soils > greywacke-derived silt loams > granular young sandy soils, and were the most variable in clayey gley loam and silt loam over gravels. Microbial reduction was 100% in allophanic soil, 16 to 18 log m−1 in pumice soil (where the unit log is the log10 reduction in maximum concentration compared with the original concentration), and was lowest in clayey gley soil (0.1–2 log m−1). For most of the other soils, the reduction was 2 to 3 log m−1, except for 9 to 10 log m−1 for fecal coliforms in a fine sandy loam. The detachment rate was only 1% of the attachment rate, indicating irreversible attachment of microbes. Soil structure (macroporosity) appeared to play the most important role in the transport of microbes and Br, while soil lithology had the greatest influence on attenuation and mass exchange. The general pattern of predicted mobile water content agrees with the measured macroporosity, which is positively related to leaching vulnerability but negatively related to dispersivity.