Global radiative forcing of nitrate and ammonium aerosols has mostly been estimated from aerosol concentrations calculated at thermodynamic equilibrium or using approximate treatments for their uptake by aerosols. In this study, a more accurate hybrid dynamical approach (HDYN) was used to simulate the uptake of nitrate and ammonium by aerosols and the interaction with tropospheric reactive nitrogen chemistry in a three-dimensional global aerosol and chemistry model, Umich/IMPACT, which also treats sulfate, sea salt and mineral dust aerosol. The calculated sulfate, ammonium and nitrate aerosol concentrations show good agreement with the available ground-based measurements over both ocean and land areas. The global annual average nitrate aerosol burden is 0.16 Tg N, with 43% (i.e., 0.079 Tg N) in the fine mode (D < 1.25 ¿m) that scatters most efficiently. The global annual average ammonium burden is 0.29 Tg N with 92% in the fine mode. A sensitivity study with a thermodynamic equilibrium model underestimates the fine-mode nitrate aerosol burden by 25%, because of the excessive nitrate formation on coarse aerosols. These underpredictions are especially important in the remote continents or over the oceans, where the availability of the total nitrate is limited. We also examined two common approaches used to treat nitrate and ammonium aerosols in global models, including the first-order gas-to-particle approximation based on uptake coefficients (UPTAKE) and a simple hybrid method that combines the former with an equilibrium model (HYB). The two methods calculate higher nitrate aerosol burdens than HDYN by +106% and +47%, respectively. Both fine- and coarse-model nitrate aerosols are overestimated by UPTAKE, but the overestimation by HYB is mainly due to uptake of nitrate by the coarse aerosols. As a result, HYB calculates lower surface concentrations of the fine-mode nitrate aerosol by up to 50% over most continental areas, compared to HDYN. Surface HNO3 and NOx concentrations are underpredicted by HYB by up to 90% and 5%, respectively. Since the reaction of N2O5 on sulfate aerosols is not included in the UPTAKE method, the NOx burden and surface concentrations are overestimated by 56% and a factor of 2--5, respectively. These results suggest the importance of using the more accurate hybrid dynamical method in the estimates of both aerosol forcing and tropospheric ozone chemistry.