This study summarises the local calibration of AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG) for flexible pavements in Tennessee. Data retrieved from the Highway Pavement Management Application of Tennessee were utilised. The alligator cracking, wheel-path longitudinal cracking, and roughness transfer functions were calibrated. The curve fitting procedure in MATLAB was used to determine the coefficients for those transfer functions. The Jackknife method was used to validate the calibrated models. Generally, MEPDG underestimated the alligator cracking for Tennessee, but overestimated the longitudinal cracking. After the calibration, both the bias and the standard error of estimate were reduced, and the predictability of those models increased appreciably. With all distress models like alligator cracking, longitudinal cracking, and rutting models being calibrated, the accuracy of the prediction from the international roughness index model increased noticeably. Jackknife is not only an unbiased way to validate the calibrated model, but also an efficient tool to detect outliers. Due to the limits in availability of data for calibrating the alligator cracking and longitudinal cracking, it is recommended that more data from different functional classes of highway, more sections with higher level of distresses and detailed traffic information be included in the future for calibration and implementation activities.