AbstractRain gardens are commonly used as green infrastructure to reduce runoff volume and introduce ecological benefits to urban areas. The objective of this study was to evaluate hydrologic performance of a rain garden constructed on low-permeability subsoil with three different drainage systems: no drain, underdrain, and surface drain. A surface-subsurface hydrologic model was implemented in the evaluation. The mathematical model incorporates processes of runoff inflow, direct precipitation, infiltration, evapotranspiration (ET), exfiltration, and outflow. This modeling assessment used a rain garden with subsoil permeability of 1 mm/h, ponding and media depth of 304.8 mm each, and loading ratio (contributory catchment area over rain garden area) of 5:1 under rainfalls over an entire year as well as individual design storms in New Jersey. Over an entire year, the longest continual ponding times are 880, 8, and 12 h for design alternatives of no drain, underdrain, and surface drain, respectively. Mosquito issues might arise from long ponding time for the no-drain design. Soil moisture of all three drainage designs never reaches down to the wilting point. Volume reduction of stormwater runoff for the three drainage designs are 89%, 15%, and 58%, respectively. The hydrologic performance for individual design storms (1, 2, 5, 10, 25, 50, and 100 year) was also addressed. Differences of effectiveness in runoff volume reduction for individual storms among the three drainage designs are consistent with those for the entire year. However, the underdrain design has the best performance in peak runoff rate reduction. Taken together, the surface drainage system has the best balance among runoff reduction for flood mitigation, soil moisture for plant survival, and ponding time for mosquito issues. The surface drainage design is recommended for a rain garden if it is installed on subsoil of low permeability and if the runoff captured does not require a substantial water quality improvement.