Many studies have reported the associations between long-term exposure to PM2.5 and increased risk of death. But, to our knowledge, none of them have used causal modeling approach or controlled for long-term temperature exposure. Few have used a general population sample. The researchers estimated the causal effects of long-term PM2.5 effect on mortality and tested the effect modifications by seasonal temperatures, census-tract-level socio-economic variables, and county-level health conditions.