The field scale experiment was performed to examine the effect of physical design parameters on the constructed wetland performance and recommend the feasible design of constructed wetland in Korean polder areas. Four sets (each set of 0.85 ha) of wetland (0.8 ha) and pond (0.08 ha) systems were used. Two different wetland systems, a wetland-pond system and a pond-wetland system, were studied to examine the effect of wetland and pond configuration. And two different length-to-width ratios were used, 2: 1 and 0.8: 1, to examine the effect of aspect ratio. A pond-wetland system was more preferable than a wetland-pond system, and also requires a smaller area than a wetland-pond system or a wetland system to reduce T-P. There was no difference in effluent concentration between the 2:1 system and the 0.8:1 system. Although the linear velocity of the 2:1 aspect was higher than the 0.8:1 aspect, resuspension was not a factor in this study due to a very low linear velocity. From this study and other literature review, it was found that design method of paddy rice field could be applied and expanded to the design of constructed wetland in Korea. Further investigation for the detailed design parameters of constructed wetland needs be continued for design method of paddy rice to be applied in full scale.