Removals of indigenous coliphage and seeded poliovirus type 1 during simulated soil-aquifer treatment were evaluated during transport of secondary effluent under unsaturated flow conditions in 1-m soil columns. Independent variables included soil type (river sand or sandy loam) and infiltration rate. Removal of coliphage was in all cases less than removal of poliovirus type 1 (strain LSc-2ab), supporting contentions that indigenous coliphage can act as a conservative indicator of groundwater contamination by viral pathogens of human origin. Coliphage retention was significantly more efficient (p<0.001) in the finer-grained sandy loam (93%) than in sand (76%). Increasing reactor detention time from 5 to 20 h increased coliphage attenuation from 70% to 99% in a 1-m sand column. There was a significant linear correlation (p=0.012) between log-transformed (fractional) coliphage concentration [log(C/C(0))] and reactor detention time. Re-mobilization of attached coliphage occurred during simulated rainfall using low-ionic-strength water. Inhibition of aerobic respiration resulted in significantly less efficient coliphage attenuation (p=0.033), suggesting the involvement of aerobic microorganisms in the survival/retention of this virus.