The purpose of this study was to identify the irrigation intervals and the amount of suitable growing substrate needed to achieve the desired shallow-extensive green roof system during a dry summer in Korea. In terms of treatment, three types (SL, P6P2L2, P4P4L2) with varying soil mixture ratios and two types (15 cm, 25 cm) with varying soil depths were created. The results have been analyzed after measuring growth and soil water contents. The difference of growth by treatment was significant in terms of green coverage, height, leaf width and photosynthesis. In measurement of chlorophyll content, no difference was detected when measured against soil depth. According to the growth measurement of Zoysia japonica with respect to differing soil mixture ratios in the 15 cm-deep treatment, a statistical difference was detected at the 0.05 significance level in photosynthesis. In case of green coverage, height, chlorophyll content and leaf width, no statistical significance was observed. In case of the 25 cm-deep treatment, a statistical significance was observed in height and photosynthesis. In terms of green coverage, chlorophyll content and leaf width, no statistical significance was detected. In comparisons of soil moisture tension and soil water contents, the irrigation interval and amount were 8 days and 14.9 L in the SL (15 cm) treatment, respectively. The irrigation interval showed for 10 days a 1.3-fold increase, and the irrigation amount was 27.4 L 1.8-fold more than SL (25 cm), respectively. For P6P2L2 (15 cm) treatment, the irrigation interval and amount were 12 days and 20.7 L, respectively. However, an irrigation interval under P6P2L2 (25 cm) was for 15 days 1.3 times longer than P6P2L2 (15 cm), and an irrigation amount of 40 L was 1.9 times more than that under P6P2L2 (15 cm). In P4P4L2 (15 cm) treatment, it was indicated that the irrigation interval was 15 days, and the irrigation amount was 19.2 L. It was not needed to irrigate for 16 days under P4P4L2 (25 cm) treatment during the dry summer and the longest no-rain periods. The irrigation interval and amount under P4P4L2 were 1.8-fold and 1.3-fold, respectively, more than SL treatment as affected by soil mixture ratio. Comparatively P4P4L2 had more 1.3-fold and 0.9-fold in irrigation interval and amount more than P6P2L2. Therefore, it can be noted that different soil depth and soil mixture ratios had a significant effect on the irrigation interval and amount.