Caffeine is a purine alkaloid naturally found in many species of plant and can be degraded by bacteria. Prolong caffeine consumption is well-known to have serious adverse effects. A caffeine-degrading bacterium had been shown to be inhibited by high concentration of caffeine. To study this phenomenon in greater detail, the effect of caffeine on the maximum specific growth rate on caffeine is studied using various inhibition kinetic models such as Luong, Haldane, Aiba, Yano, Teissier, Webb and Monod. The Haldane model was the best model in modelling the effect of caffeine concentration to Caulobacter crescentus maximum specific growth rate based on statistical tests such as corrected AICCc (akaike information criterion), bias factor (BF), adjusted coefficient of determination (R2) and root-mean-square error (RMSE). Parameters obtained from the fitting exercise were maximum specific growth rate (qmax), Ks (concentration of substrate at the half maximal specific growth rate (mg/L) and Ki (inhibition constant) mg/L with their values at 0.7620 (95%, C.I. or confidence interval from 0.603 to 0.920), 0.1050 mM (95% C.I. from 0.001 to 0.211) and 4.614 mM (95% C.I. from 2.154 to 7.073), respectively. These biologically meaningful coefficients will be useful in predicting growth conditions under batch studies and also probably under actual bioremediation strategy where the concentration of caffeine in the treatment may need be diluted to nontoxic concentration prior to remediation.