Soil gas 220Rn and 222Rn concentrations were measured at sites in Chester and Aberdeen, New Jersey (NJ). Two years of 222Rn and 220Rn data were obtained from a depth of 0.85 m, followed by two years from depths of 0.28, 0.56, 0.85, and 1.28 m. Variations of 220Rn and 222Rn, before the first winter that sample tubes were installed, were larger than later, indicating that the soil structure, disturbed during installation of the tubes, may have significantly redistributed after the first winter, thus ensuring that the sample was drawn from near the bottom of the tube. At the Chester site, autumn 222Rn concentrations were found to be up to 10 times higher than winter values, variations larger than predicted assuming diffusion-only transport. Spatial variations up to an order of magnitude were observed over distances of a few meters. The 220Rn concentrations were typically ∼2 to 3 times higher during summer than during winter. At the Aberdeen site, 220Rn and 222Rn concentrations were about an order of magnitude less than the lowest Chester site values with no statistically significant temporal or spatial variations. Permeability, thought to be an indicator of parameters controlling soil gas Rn variations, showed poor correlations with 220Rn or 222Rn at both sites.