Granular activated carbon (GAC) breakthrough predictive models were developed for 2-methylisoborneol (MIB) based on the 35 rapid small-scale column tests. The models were designed to predict the bed volumes treated (time) when the effluent concentration would reach predetermined MIB concentrations, for example, a target effluent MIB concentration of 10 ng/L. The models were created, using multiple linear regression techniques in stepwise manner to determine significant predictors of breakthrough to the target effluent conditions. Models were first developed for fresh 12x40 mesh size GAC at 10 and 20 minutes EBCT and then modified to account for 8x20 and 8x30 mesh size GAC and GAC age. The model input includes influent total organic carbon (TOC), influent MIB concentration, GAC particle size, and GAC age. The impact of intermittent loading of MIB was simulated in the model to determine the GAC bed life. Includes 8 references, figures.