Rice is one of the most important staple foods and two-third of the world’s peoples are dependent to it. Hence, reducing losses during milling process of rough rice is of great importance. Rice is usually harvested at moisture content of 25% to 35% (w.b.). In order to storage and prevent its deterioration the level of moisture content is decreased to 10% to 13% (w.b.). The most important index regarding rice milling quality is head rice yield. The main factors affecting this index are moisture gradients and finally internal cracks created during the drying process. In Iran, traditional and conventional drying methods of rough rice cause considerable losses due to non-uniformity in the process. One method to eliminate moisture gradients, and consequently reduce the losses during the drying process is intermittent drying method. In this method, after each drying pass the rough rice is located in an isolated place with specific temperature in order to balance the moisture content of different layers through preventing moisture exchange with the environment. This operation is known as tempering process. On the other hand, non-uniformity at various depths of rough rice during fixed bed drying method increases the milling losses. Because of passing the airflow through the rough rice kernels in fluidized bed drying method, the moisture content decreases uniformly in this method. This study was aimed to investigate the effect of rough rice drying in fluidized bed method with intermittent drying pattern on cracking of Hashemi (long-grain) and Koohsar (medium- grain) rough rice varieties. A laboratory vibrated fluidizes bed dryer located in a walk-in chamber equipped with a humidity control system was used to dry the samples. The drying experiments were carried out at environment relative humidity of 60% and minimum fluidization velocities of 2.4 and 2.6 m s -1 for Koohsar and Hashemi varieties in two replications. The intermittent experiments were conducted at temperatures of 60 and 80 °C, moisture reduction levels of 1.5%, 3% and 4.5%, and tempering durations of 0 (no tempering), 40, 80, 120, 160 and 200 min. By using the experimental results and moisture content analysis obtained by spherical geometry analytical model, the appropriate temperature and moisture reduction level for the drying stage, and the suitable duration for the tempering stage were determined. The numbers of internal cracks immediately and 48 h after each experiment were obtained, and then were analyzed by the stress cracking index (SCI). The results indicated that the model could acceptably predict the drying kinetics for intermittent fluidized bed drying including the kernel surface moisture content, the kernel average moisture content, and the appropriate duration for tempering process. It was also revealed that the tempering process was completed more rapidly at 80 °C. The appropriate tempering duration was obtained to be 120 min for two varieties. An increase in temperature from 60 to 80 °C had no significant effect on the SCI, however the drying duration decreased by 33.6 %. Therefore, based on both the drying duration and the SCI, the best moisture reduction level was determined to be 3% for each drying pass. Compared to the fixed bed mode (Ghasemi, 2013), under the fluidized bed method, the SCI decreased by 33.82% and 45.46% immediately after drying, and 48 h after drying, respectively. Keywords: intermittent drying, moisture gradients, spherical geometry model, stress cracking index, tempering