The two-region model was developed originally to describe nonsorbing chemical transport in soils with dead-end pores based on the concept of mobile and immobile regions in the soil. It has been shown that the model can simulate solute transport in soils with local stratification, or inhomogeneity, of hydraulic conductivity. However, the physical basis of the model becomes questionable, since the mobile-immobile region concept does not apply in stratified soils. In both soil types the nonequilibrium effect is caused by an apparent mass transfer process within the soil, as distinct from advection and diffusion. Where there are immobile regions, the mass transfer is due to solute interregion diffusion alone. In stratified soils the nonequilibrium mass transfer process is affected also by local flow variations. A conceptual model, numerical simulations, and laboratory experiments are presented to analyze these effects. For a given soil with fixed local stratification of hydraulic conductivity, it is shown that in the low- velocity range, the apparent mass transfer rate parameter, α, scales as V 2/D (V is pore water velocity in the two-region model and D is the longitudinal dispersion coefficient), which implies that the mass transfer process is predominantly affected by local flow variations. When the velocity is relatively high, α ∞ DT /h 2 (DT is the interregion diffusion coefficient and h is the characteristic thickness of the stratified layers) and the mass transfer process is dominated by interregion diffusion. These scaling relations for α reflect the two mechanisms controlling the mass transfer process in locally stratified soils. They have implications for scaling of time-dependent mass transfer from laboratory models to prototype soils. In particular, the relationship α ∞ V^2/D leads to the conclusion that exact physical modeling of nonsorbing chemical transport coupled with apparent mass transfer in locally stratified soils may be viable.