Abstract Condensation of gases (H2SO4, NH3, HNO3, low vapor pressure organics, etc.) onto existing aerosols can account for a significant portion of fine particulate matter. Since aerosols affect human health, visibility, and climate change, it is important to be able to model condensation/evaporation accurately in order to predict how particle mass will change over time. (Atmos. Environ. 34 (2000) 2957) proposed adapting the trajectory-grid method, used for solving the transport equation, to solve the condensation/evaporation equation. Their preliminary results showed it to be fast and accurate in simple systems (one component, etc.). The approach of Chock and Winkler has been modified for implementation in both the Hybrid method (Atmos. Environ. 34 (2000) 3617) and the improved multicomponent aerosol dynamics model (MADM) (Aerosol Sci. Technol. 32 (2000) 482) The first improvement in MADM modifies the method for restricting the acidic flux, while the second reduces physically meaningless dry/wet oscillations in the aerosol phase by assuming the aerosol is metastable when these oscillations are present. Measurements in Claremont, CA in August of 1987 are used to evaluate the new method in a one-dimensional model and the October 1995 PM episode in the South Coast Air Basin in CA is used for evaluation in a three-dimensional chemical transport model (PMCAMx). The trajectory-grid method allows the use of a simple scheme for time step selection and provides at least a factor of two or three reduction in computational requirements compared to an ODE solver at the same level of accuracy. The improvements to MADM also have a significant effect on performance, providing over an order of magnitude reduction in computational requirements compared to the original MADM. In the three-dimensional chemical transport model, the Hybrid method of (Atmos. Environ. (2000) 3617) is applied which assumes the smallest particles are in equilibrium while the condensation/evaporation equation is solved for the larger ones. Combined with the improvements to MADM and trajectory-grid method, this Hybrid approach takes just three to four times the computational requirements of assuming bulk equilibrium for all particles, while providing more accurate predictions of the aerosol size distribution.