Atmosphere-surface exchange fluxes can vary significantly over a large grid cell of atmospheric models and need to be estimated with subgrid-scale variations included. The present study aims at developing a method for parameterization of subgrid-scale (PASS) surface fluxes in atmospheric models, with emphasis on a scheme to distribute the mean atmospheric forcing from the model grid (MG) to subgrid (SG) cells and on using spectral vegetation indices derived from satellite remote sensing to estimate surface parameters at SG scales. The mean atmospheric forcing, including wind speed, temperature, and humidity, is distributed to SG cells according to spatial differences in these parameters at the surface and the strength of vertical turbulence mixing in the atmospheric boundary layer. Simulation with this scheme allows a feedback of changed SG atmospheric conditions caused by surface heterogeneity to influence the SG surface exchange. The SG surface exchange is parameterized with a suite of simplified relationships developed to infer surface conductance, aerodynamic resistance, surface albedo, and soil heat flux ratio on the basis of spectral vegetation indices available from satellite remote sensing, while empirical coefficients used in the relationships are derived for selected surface types from available field experiments. The computation of the surface flux and energy budget for a large number of SG cells is made feasible by using a rapid scheme based on an analytical solution to the surface energy budget equations. The PASS model was tested in a single-cell mode with field measurements and evaluated in a multicell mode with limited data and a sensitivity study. ¿ American Geophysical Union 1995