Concentrations of both fine particulate matter (PM 2.5 ) and ozone (O 3 ) in the United States (US) have decreased significantly since 1990, mainly because of air quality regulations. These air pollutants are associated with premature death. Here we quantify the annual mortality burdens from PM 2.5 and O 3 in the US from 1990 to 2010, estimate trends and inter-annual variability, and evaluate the contributions to those trends from changes in pollutant concentrations, population, and baseline mortality rates. We use a fine-resolution (36 km) self-consistent 21-year simulation of air pollutant concentrations in the US from 1990 to 2010, a health impact function, and annual county-level population and baseline mortality rate estimates. From 1990 to 2010, the modeled population-weighted annual PM 2.5 decreased by 39 %, and summertime (April to September) 1hr average daily maximum O 3 decreased by 9 % from 1990 to 2010. The PM 2.5 -related mortality burden from ischemic heart disease, chronic obstructive pulmonary disease, lung cancer, and stroke, steadily decreased by 53 % from 123,700 deaths yr −1 (95 % confidence interval, 70,800–178,100) in 1990 to 58,600 deaths −1 (24,900–98,500) in 2010. The PM 2.5 -related mortality burden would have decreased by only 24 % from 1990 to 2010 if the PM 2.5 concentrations had stayed at the 1990 level, due to decreases in baseline mortality rates for major diseases affected by PM 2.5 . The mortality burden associated with O 3 from chronic respiratory disease increased by 13 % from 10,900 deaths −1 (3,700–17,500) in 1990 to 12,300 deaths −1 (4,100–19,800) in 2010, mainly caused by increases in the baseline mortality rates and population, despite decreases in O 3 concentration. The O 3 -related mortality burden would have increased by 55 % from 1990 to 2010 if the O 3 concentrations had stayed at the 1990 level. The detrended annual O 3 mortality burden has larger inter-annual variability (coefficient of variation of 12 %) than the PM 2.5 -related burden (4 %), mainly from the inter-annual variation of O 3 concentration. We conclude that air quality improvements have significantly decreased the mortality burden, avoiding roughly 35,800 (38 %) PM 2.5 -related deaths and 4,600 (27 %) O 3 -related deaths in 2010, compared to the case if air quality had stayed at 1990 levels.