Permeable pavement (PP) can be used to decrease urban surface runoff. However, few studies have been conducted to explore the runoff reduction effect of various structures of PP in the carriageway. In this study, several structures of PP used in the carriageway of sponge cities in China were investigated and divided into three types: surface drainage, base course storage and drainage, and fully permeable. Then, the runoff models were developed by Storm Water Management Model to simulate the effect of the three types under various rainfall recurrence periods. Results show that rainfall recurrence period, structure and thickness of the permeable layer were identified as the most influential factors in PP runoff reduction. The surface drainage can reduce total runoff depth and coefficient by more than 14%, and also delay runoff start time and duration by more than 40 minutes. Surface runoff in the base course storage and drainage can only be generated when recurrence period is 50 years. The fully permeable does not generate any runoff under all recurrence periods. Based on simulation results, a series of runoff coefficient values for PP were recommended to help the design and implementation of PP in mitigating urban waterlogging problems.