The presence of air-water interface in batch sorption experiments may result in inaccurate estimation of virus adsorption onto various soils. A batch sorption experiment was conducted to compare the adsorption results of MS2 in different soils under presence/absence of air-water interface. Soils with sterilization/nonterilization treatment were used. Virus recovery efficiency in a blank experiment (no soil) was also evaluated as affected by different amount of air-water interface. The presence of air-water interface altered the results of virus adsorption in different soils with different extent, with Sandy fluvo-aquic soil being the most considerably affected, followed by Red loam soil, and the least being Red clay soil, probably because of different soil properties associated with virus adsorption/inactivation. Soil sterilization resulted in more significant difference of virus adsorption onto the Sandy fluvo-aquic soil between the presence and absence of air-water interface, while a reduced difference was observed in the Red loam soil. The presence of air-water interface significantly decreased virus recovery efficiency, with the values being decreased with increase in the amount of air-water interface. Soil particles likely prohibit viruses from reaching the air-water interface or alter the forces at the solid-water-air interface so that the results from the blank experiment did not truly represent results from control blank, which probably resulted in adsorption difference between presence and absence of the air-water interface.