Repeated load triaxial (RLT) tests were conducted on the granular and subgrade materials used at CAPTIF (NZ Transport Agency's test track). Permanent strain relationships found from RLT testing were later used in finite element models to predict rutting behaviour and magnitude for the pavements tested at the CAPTIF test track. Predicted rutting behaviour and magnitude were close to actual rut depth measurements made during full-scale pavement tests to validate the methods used. This method of assessing rutting in granular materials was used to predict the life or number of axle passes to achieve a rut depth defining the end of life for a range of pavement thicknesses, and the subgrade types to produce new pavement thickness design charts. The results of these rut depth predictions showed the Austroads guide required thicker pavements for low traffic volumes, while the rut depth predictions showed significantly thicker pavements were required for high traffic volumes. In fact the rut depth predictions indicated the traffic loading limits for granular pavements were around 7 million equivalent standard axle passes (ESAs) for the subgrade California bearing ratio (CBR) 2% and 11 million ESAs for the subgrade CBR 8%.