AbstractLittle field‐scale research has been done to evaluate the effectiveness of compacted soil barriers in retarding the movement of water and leachates. In response to this need, the Illinois State Geological Survey constructed and instrumented an experimental compacted soil liner. Infiltration of water into the liner has been monitored for two years. The objectives of this investigation were to determine whether a soil liner could be constructed to meet the U.S. EPA's requirement for a saturated hydraulic conductivity of less than or equal to 1.0×10−7 cm/s, to quantify the areal variability of the hydraulic properties of the liner, and to determine the transit time for water and tracers through the liner.The liner measures 8 m×15 m×0.9 m and was designed and constructed to simulate compacted soil liners built at waste disposal facilities. The surface of the liner was flooded to form a pond on April 12, 1988. Since flooding, infiltration has been monitored with four large‐ring (LR) and 32 small‐ring (SR) infiltrometers, and a water‐balance (WB) method that accounted for total infiltration and evaporation. Ring‐infiltrometer and WB data were analyzed using cumulative‐infiltration curves to determine infiltration fluxes. The SR data are lognormally distributed, and the SR and LR data form two statistically distinct populations. Small‐ring data are nearly identical with WB data; because there is evidence of leakage in the LRs, the SR and WB data are considered more reliable.Geostatistical analysis of the SR infiltration data revealed that the infiltration‐flux data were unstructured (random) at scales greater than 0.8 m. This analysis shows that it is possible to construct a compacted soil liner with a uniformly low saturated hydraulic conductivity, and that classical statistics should adequately estimate the mean infiltration flux of the liner and the associated uncertainty in that value.Saturated hydraulic conductivity of the liner was estimated using Darcy's Law and the Green‐Ampt Approximation; the average values for these calculations, based on the first and second years of SR data, were 4.0×10−8 and 3.4×10−8 cm/s, respectively. Breakthrough of water at the liner's bottom is expected to occur approximately six years after the initial ponding of the liner.