Natural ventilation is an energy-efficient design approach to reduce infection risk (IR), but its optimized design in a coach bus environment is less studied. Based on a COVID-19 outbreak in a bus in Hunan, China, the indoor-outdoor coupled CFD modeling approach is adopted to comprehensively explore how optimized bus natural ventilation (e.g., opening/closing status of front/middle/rear windows (FW/MW/RW)) and ceiling wind catcher (WCH) affect the dispersion of pathogen-laden droplets (tracer gas, 5 μm, 50 μm) and IR. Other key influential factors including bus speed, infector's location, and ambient temperature (Tref) are also considered. Buses have unique natural ventilation airflow patterns: from bus rear to front, and air change rate per hour (ACH) increases linearly with bus speed. When driving at 60 km/h, ACH is only 6.14 h−1 and intake fractions of tracer gas (IFg) and 5 μm droplets (IFd) are up to 3372 ppm and 1394 ppm with ventilation through leakages on skylights and no windows open. When FW and RW are both open, ACH increases by 43.5 times to 267.50 h−1, and IFg and IFd drop rapidly by 1–2 orders of magnitude compared to when no windows are open. Utilizing a wind catcher and opening front windows significantly increases ACH (up to 8.8 times) and reduces IF (5–30 times) compared to only opening front windows. When the infector locates at the bus front with FW open, IFg and IFd of all passengers are <10 ppm. More droplets suspend and further spread in a higher Tref environment. It is recommended to open two pairs of windows or open front windows and utilize the wind catcher to reduce IR in coach buses.