Constructed wetlands are typical best management practices (BMPs) often used to reduce nutrient loads in streams. Evaluating the effectiveness of wetland design on nutrient removal is essential to assist watershed managers in optimal design of BMP dimensions and placement. In this study, we assess performance of two constructed wetland systems (comprising of nine wetlands) installed in downstream of Longhongjian Stream in Hangzhou City, China. These wetland systems are monitored and evaluated for their effects on nutrient removal, particularly TN, NO3-N, TP, and PO4-P. Based on wetland input–output metrics, removal efficiency (RE) is used to quantify wetland system. Results show that both wetland systems effectively removed nutrients, with RE as high as 45% of TN, 57% of NO3-N, 78% of TP, and 86% of PO4-P. In general, nutrient removal efficiency is seasonally dependent, with better removal efficiency occurring during warmer seasons than others. Macrophyte uptake is a primary removal process in these wetlands. We observe that more wetlands working concurrently can provide a greater level of control on nutrients in lotic environments. Wetland design parameters play an important role in removal of nutrients in streams. Increasing flow volume and surface area of wetland, designing curvilinear shoreline, and longer flow paths can be used as design criteria for wetland systems aimed at nutrient removal.