The mitigative impact of covering the floor of a crawl space with a membrane has been studied under welldefined and controlled conditions. The measurements have been done with a homogeneous column of dry sand covered with a sheet of polyethylene foil. An airfilled volume on top of the column simulates a crawl space. The experiments mainly concern long-term measurements of the crawl-space radon concentration in combination with steady-state diffusive and combined advective and diffusive transport through the sand column and crawl space. The experimental data are analyzed with both simplified mass-balance models for radon entry into the crawl space and with a two-dimensional numerical model based on a finite-difference approach. In all experiments the influence of atmospheric pressure variations is clearly present. For most experiments the agreement between calculations, which make use of independently measured transport parameters for both sand and foil, and measurements is within 10%. However, the discrepancy is larger for experiments with continuous advective transport from the crawl space to the sand. With undamaged foil, the calculations overestimate the measurements by 20%. Reversely, with an opening in the center of the foil, the calculations underestimate the measurements by 20–40%. The results show that under controlled conditions radon transport from a foil-covered sand column into a crawl-space can be described within 40% on basis of separately measured parameters.