Contaminant transport through a single hole in a geomembrane (GM) that forms part of composite liner with a geosynthetic clay liner (GCL) for applied heads (0.3 and 1 m) and an applied pressure of about 100 kPa is modeled and compared with experimental results. The system was first permeated with distilled water until a steady state was reached. The permeant was then switched to an NaCl solution. The inferred interface transmissivity at steady state (after 280 days) is about 2.3 × 10−11 m2/s for both heads examined. After 800 days permeation with an 0.14 mol/L NaCl solution, the interface transmissivity between the geomembrane and GCL was inferred to have decreased from 2.3 × 10−11 to 1.1 × 10−11 m2/s for experiments with heads of both 0.3 and 1 m. Modeling of the flow and changes in the GCL hydraulic conductivity arising from interaction between the bentonite and the 0.14 mol/L NaCl permeant is shown to give excellent agreement with the observed leakage. The differences between modeling a uniform and nonuniform interface transmissivity is found to be small — with the leakage being dominated by the interface transmissivity 0.1 to 0.15 m from the hole in the GM. Typical published values for the diffusion coefficient for both the GCL and sandy soil are shown to give good predictions of effluent concentration with time and chloride distribution in the silty sand in contact with the composite liners at the end of the test.