Although the temporal-scale effects on PM_2.5–meteorology associations have been discussed, no quantitative evidence has proved this before. Based on rare 3 h meteorology data, we revealed that the dominant meteorological factor for PM_2.5 concentrations across China extracted at the 3 h and 24 h scales presented large variations. This research suggests that data sources of different temporal scales should be comprehensively considered for better attribution and prevention of airborne pollution.