n recent years, due to the awareness of quality, many advances have been occurred in quality assessment techniques of fruits, vegetables and food. Meanwhile, laser light back-scattering imaging (LLBI) method is regarded as an efficient and non-destructive technique for predicting the quality indices of agricultural materials. On the other hand, drying is one of the most important postharvest activity for processing the agricultural products. Determining the moisture content of the product during drying is an extremely important operation to control the product's condition and evaluate the drying performance. The conventional measurement method involves placing the product on the scale, which is due to the contacting nature, this method has its own limitations. In this research, we tried to measure the moisture content of the quince during the drying process using LLBI method. Quince is considered as a food-medicinal fruit since it contains combination of pectin and high fiber. In order to dry the quince slices and simultaneously acquire their laser light back-scattering images, a special drying set-up was first designed and constructed. This set-up was then connected to a hot air dryer with the adjustable temperature and speed. Using a digital camera placed on the top of the set-up, the light back-scattered images were collected by illuminating the infrared (wavelength of 980 nm with 25 mW power) and green (wavelength of 532 nm with 1 mW power) lasers. All experiments were carried out at a constant speed of 0.5 m/s. The drying temperature was adjusted to three levels of 50, 60, and 70 ? C. After determining the initial moisture content of the samples using the standard oven method, the remaining slices were then used for drying and imaging purposes. The experiments were performed over a period of 3 days, and in each day, the laser light back-scattering images of 10 quince samples were acquired and the samples were immediately weighed. This was repeated every 30 minutes until the samples reached to a constant weight. In order to analyze the images, the region of back-scattered light were segmented from the background using the Mahalanobis distance method. Numerical histogram and statistical variables of the back-scattered images were selected as the input features. Also, for each image, a plot obtained from radial distance of saturation region from its center of gravity against the rotation angle was obtained using the Canny method. Finally, the combination of histogram feature and the saturation region plot was evaluated to predict the moisture content. Partial least squares (PLS) method was used to extract the predictive models for estimating the moisture content of quince slices during drying. The results showed that there was a direct correlation between the variance of the saturation plot and the moisture content, where by decreasing the moisture content, the variance in the saturation region decreased. Moreover, there was a direct correlation between the amount of back-scattered light and the moisture content, where by decreasing moisture content, the area of the back-scattered region decreased. Comparison of the ability of both infrared and green lasers as well as the combination of these two lasers in two steps of calibration and cross-validation showed that the best results obtained from combination of two lasers. However, when the data of a single laser was used, the green laser had a better performance with respect to infrared one. Among the different lasers and features, the best PLS models obtained from combination of two lasers and histogram-saturation plot as the input feature. In test set validation stage, the best model could predict the moisture content of quince slices with a coefficient of determination ( R 2 p ) of 0.855, root mean squared error of prediction (RMSEP) of 10.020%, and residual predictive deviation (RPD) of 2.61.