In this study, the uncertainty during sample processing or homogenization of several different matrices (tomato, lettuce and orange) was determined at ambient temperature and with dry ice. The method was based on surface treatment of the matrix with a reference compound (C-chlorpyrifos) and estimation of the sampling constants. The uncertainty determination was carried out developing 5 experiments for each of the different homogenizations and matrices. After homogenization, 5 replicate analytical portions of 15 g and 150 g were withdrawn for each experiment. The variance was calculated for each analytical portion. The significant difference between the variance of smaller analytical portion and the variance of larger analytical portion was evaluated. Then the coefficient of variation (CV) was calculated for the larger portion. This coefficient was used to calculate the sampling constant. Then, the uncertainty could be determined for any analytical portion size. The sample processing uncertainty obtained with dry ice for all samples was not significantly different. There was no any significant difference between homogenization at ambient temperature and homogenization with dry ice for each matrix. The extraction uncertainty was negligible for all matrices compared to the sample processing component. It allows one to quantify the sample processing uncertainty directly.