The mechanism and environmental controls on soil hydrogen (H2) uptake are not well understood but are essential for understanding the atmospheric H2 budget. Field observations of soil H2 uptake are limited, and here we present the results from a series of measurements in forest, desert, and marsh ecosystems in southern California. We measured soil H2 fluxes using flux chambers from September 2004 to July 2005. Mean H2 flux rates and standard deviations were −7.9 + −4.2, −7.6 + −5.3 and −7.5 + −3.4 nmol m−2 s−1 for the forest, desert, and marsh, respectively (corresponding to deposition velocities of 0.063 + −0.029, 0.051 + −0.036, 0.035 + −0.013 cm s−1). Soil profile measurements showed that H2 mixing ratios were between 3% and 51% of atmospheric levels at 10 cm and that the penetration of H2 into deeper soil layers increased with soil drying. Soil removal experiments in the forest demonstrated that the litter layer did not actively consume H2, the removal of this layer increased uptake by deeper soil layers, and the exposure of subsurface soil layers to ambient atmospheric H2 levels substantially increased their rate of uptake. Similar soil removal experiments at the desert site showed that extremely dry surface soils did not consume H2 and that fluxes at the surface increased when these inactive layers were removed. We present a model of soil H2 fluxes and show that the diffusivity of soils, along with the vertical distribution of layers that actively consume H2 regulate surface fluxes. We found that soil organic matter, CO2 fluxes, and ecosystem type were not strong controllers of H2 uptake. Our experiments highlight H2 diffusion into soils as an important limit on fluxes and that minimum moisture level is needed to initiate microbial uptake.