COVID-19 has prematurely ended many lives, particularly among the oldest Americans, but the pandemic has also had an indirect effect on health and non-COVID mortality among the working-age population, who have suffered the brunt of the economic consequences. This analysis quantifies the changes in mortality for selected causes of death during the COVID 19 pandemic up to December 31, 2020, and investigates whether the levels of excess mortality varied by age group. The data comprise national-level monthly death counts by age group and selected causes of death from January 1999 to December 2020 combined with annual mid-year population estimates over the same period. A negative binomial regression model was used to estimate monthly cause-specific excess mortality during 2020 controlling for the pre-pandemic mortality patterns by age, calendar year, and season. To determine whether excess mortality varied by age, we tested interactions between broad age groups and dichotomous indicators for the pre-pandemic (January-February) and the pandemic (March-December) portions of 2020. In relative terms, excess all cause mortality (including COVID-19) peaked in December at ages 25-44 (RR=1.58 relative to 2019, 95% CI=1.50-1.68). Excluding COVID-19, all of the excess mortality occurred between ages 15 and 64, peaking in July among those aged 25-44 (RR=1.45, 95% CI 1.37-1.53). We find notable excess mortality during March-December 2020 for many causes (i.e., influenza/pneumonia, other respiratory diseases, diabetes, heart disease, cerebrovascular disease, kidney disease, and external causes), but almost exclusively among young and midlife (aged 25-74) Americans. For those aged 75 and older, there was little excess mortality from causes other than COVID-19 except from Alzheimer's disease. Excess non-COVID mortality may have resulted, at least partly, from incorrectly classified COVID-19 deaths, but neither misclassification nor an atypical flu season that disproportionately affected younger people is likely to explain the increase in mortality from external causes, which was evident even during January-February 2020. Exploratory analyses suggest that drug-related mortality may be driving the early rise in external mortality. The growth in drug overdoses well before there was any hint of a pandemic suggests it is probably not solely an indirect effect of COVID-19, although the pandemic may have exacerbated the problem.