Green roofs are one type of Low Impact Development (LID) technique used to retain stormwater. Soil depth and slope of green roofs are important factors when considering their effectiveness. LIDs designs are usually based on the estimation of runoff depths, volumes, and peak flows estimated by simple methods such as the Rational Method and the National Resources Conservation Service (USDA-NRCS) Curve Number (CN). Experimental rainfall-runoff data was used, from a set of green roofs with two different slopes (2% & 33%) and two different depths of soil (4-in & 6-in) with a conventional roof surface, to determine representative values of CN estimated for different initial abstraction rations (0, 0.05, 0.2). CN values ranged from 84.66 to 94.29, but no significant differences (p < 0.05) were observed among depth-slope treatments or initial abstraction ratios. The higher CN values were observed for the initial abstraction ratio of 0.2 (90.45 - 94.29) despite the slope and soil depth combinations. Green roofs with 4-inch depth had higher CN values (90.65 - 94.29) than those with 6-inch depth (84.66 - 93.60). Similarly, the CN was higher for structures with a 33% slope (90.35 - 94.29). The goodness of fit to predict rainfall excess for each treatment by the Curve Number method (0.71
