Abstract: Abstract: The drying process is one of the most important processes in food engineering. Heat and moisture transfer in fruits and vegetables during drying is a complex process, and knowledge of the moisture profile is fundamentally important for industrial processes, because the quality aspects of dried foodstuffs, such as nutrient content, safety and weight, are related to moisture content. A better understanding of the mechanism of moisture transfer should help improve product quality and the efficiency of the drying process. Such information will also help produce realistic computer simulations of drying processes and Fickian and non-Fickian models of moisture transport. Moisture profiles in foodstuffs can be measured using destructive and non-destructive methods. Slicing and freezing is not an accurate method because of low precision and moisture loss during cutting process, although it could give inside moisture distribution. In recent years, the applicability of nuclear magnetic resonance imaging (MRI) to measure mass transport phenomena in porous systems, especially in foodstuffs and biological materials, including drying processes, has been demonstrated. NMR imaging as a non-destructive, non-invasive, promising technique has been used to obtain moisture profiles during drying of vegetables and fruits. In this paper, the moisture transport in a cylindrical carrot sample was visualized and studied using nuclear magnetic resonance images obtained from the drying experiments. The transient moisture profile distributions in carrot were measured during the hot-air drying process with air temperatures of 40 and 70℃, respectively. Experimental results showed that the moisture profiles moved irregularly toward the center of the carrot sample in both the axial and radial directions, indicating a muti-dimensional and unsteady-state mass transfer process that has non-Fickian moisture transport characteristics. In the initial drying stage, a sharp moisture gradient was found indicating significant moisture flux at the surface of the carrot. With the process of drying, the ratio of the MRI diameter declined faster than that of the optical diameter, which indicated that the dried layer appeared at the surface and moisture profile moved inside. The Henderson-Pabis model (MR=1.003e-0.01114t, R2=0.9994) achieved better predictive accuracy than other models and satisfactorily described drying characteristics of the carrot cylinder at 70℃. The maximum relative error of prediction compared with the measured results was 7.69%, with relative errors during the drying process at 70℃ commonly remaining less than 4%. The moisture transport of the carrot center layer was simulated by the Henderson-Pabis model (MR=1.005e-0.00286t，R2=0.9978) during the drying process at 40℃.These results could assist in the optimization of drying process and theoretical simulation on moisture transport considering shrinkage caused by drying.