Husk separator is an indispensable equipment in rice milling plants. However, any basic research on the designing and operating criteria of the husk separator have rarely been conducted in Korea. According to the survey results reported recently, grain loss occurs in the process of rice husk separation at custom rice milling plants in Korea and the performance of husk separator has also not been identified. With this regard this study was conducted with a typical commercial husk separator to investigate the effect of the operational factors such as feed rate, blower speed and opening ratio on the velocity distribution in the air duct and the performance of the separator. The results are summerized as follows: 1. The average wind velocity in the primary air duct increased linearly with the blower rpm and the size of air inlet port in both cases of double type and single type operations. 2. The coefficient of variation in the horizontal wind velocities in the primary air duct was the minimum when the opening ratio was 0.22 ($0.052m^2$ of air inlet port) in both cases of single type and double type operations regardless of the blower speeds used in this test. The average wind velocity at the upper part of air duct was greater by 2-5 m/s than the velocity at the bottom part in double type operation. In case of single type operation, however, the average velocity in the middle part was greater than the upper or bottom part when the opening ratio was greater than 0.74. 3. The relationship between the overall effectiveness of separation(Ed for double type and Es for single type) and the average wind velocity (Va) in the primary air duct was expressed in the following quadratic functions. $$Ed=-190.84+106.18Va-10.052Va^2$$ ($r^2$ = 0.97782) $$Es=-223.76+106.23Va-9.1935Va^2$$ ($r^2$ = 0.97029) The average wind velocity required to obtain the overall effectiveness of separation more than 80% ranged from 4.04 m/sec to 5.84 m/sec in case of double type operation, and from 4.70 m/sec to 6.20 m/sec in case of single type. 4. An optimum wind velocity can be obtained with an increase in the blower speed or the size of air inlet port as presented in Figure 8. There was a tendency that the faster the blower speed, the narrower the control range of the air inlet port. 5. The feed rates (1850kg/hr and 2100kg/hr) adopted in this experiment did not bring about a significant difference in both the overall effectiveness of separation and the power consumption. 6. The energy consumption increased cubically with the blower speed but linearly with the size of the air inlet port. On the basis of the results described in items 1, 3, and 6, it would be more economic to adjust the size of the air inlet port larger with a relatively low blower speed than to adjust the size smaller with a relatively high speed.