A negative binomial regression model was used to study the association between 3-year average PM2.5 concentrations and age-adjusted mortality rates for 105 of the 345 municipalities in Chile. Models were fitted for all (ICD10 A to Q codes), cardiopulmonary (I and J), cardiovascular (I), pulmonary (J), cancer (C), and lung cancer (C33-C34) causes; controlling for meteorological, socioeconomic, and demographic characteristics. A significant association of PM2.5 exposure with cardiopulmonary (relative risk for 10 µg/m3 PM2.5: 1.06; 95% confidence interval = 1.00, 1.13) and pulmonary (1.11; 1.02, 1.20) age-adjusted mortality rates was found. Cardiovascular (1.06; 0.99, 1.13) and all causes (1.02; 0.98, 1.07) were positive, but not significant. No significant association was found between cancer and lung cancer. The positive associations remained even when controlling for multiple confounding factors, model specifications, and when considering different methods for exposure characterization. These estimates are in line with results from cohort studies from the United States and European studies. Three-year average PM2.5 exposure is positively associated with the age-adjusted mortality rate for cardiopulmonary and cardiovascular causes in Chile. This provides evidence of the medium-term exposure effect of fine particles on long-term mortality rates.