The results of a diffusion modelling study to evaluate experimental data on oxidative induction time (OIT) depletion from high-density polyethylene (HDPE) geomembrane (GM) in accelerated ageing tests are presented. The paper provides: (1) results of diffusion modelling of OIT depletion from a GM immersed in leachate and in a composite liner with leachate above the liner at different incubation temperatures; (2) a comparison of the results of the diffusion model and the conventional first-order (exponential) antioxidant depletion model; (3) estimates of diffusion and partitioning coefficients at typical landfill temperatures based on Arrhenius-type relationships; and (4) an application of the estimated parameters to model a composite liner with 30 cm thick sand layer. The antioxidant diffusion coefficients ranged from 2.1 × 10−15 (at 26°C) to 1.6 × 10−13 m2/s (at 85°C) and the partitioning coefficients ranged from 720 (at 26°C) to 4 (at 85°C). The antioxidant depletion time obtained using the first-order model was similar to that predicted using the diffusion model for tests where the OIT was depleted during the test period. However the first-order model gave smaller predictions of depletion time than the diffusion model in cases where there was only limited OIT depletion and in these cases the diffusion model is likely to the give more accurate predictions. Arrhenius modelling provided a means of estimating diffusion and partitioning coefficients at field temperatures. At a typical landfill temperature of 35°C the calculated antioxidant depletion time for the geomembrane considered was about 130 years for a case where there was a 1.5 cm sand protection layer and 230 years for the case when 30 cm sand protection layer was used. Thus these results suggest that the use of a 30 cm sand protection layer in addition to the typical geotextile protection layer between the geomembrane and a coarser granular leachate drainage layer would provide potential benefits in terms of extending the geomembrane service life by reducing the rate of outward diffusion of antioxidants from the geomembrane (as well as providing good physical protection of the liner). This paper has also illustrated how diffusion modelling can be used for considering a range of situations different from those under which the basic experimental data was obtained.