Abstract.A new modal aerosol parameterization scheme, the statistical-numerical aerosol parameterization (SNAP), was developed for studying aerosol processes and aerosolcloud interactions in regional or global models. SNAP applies statistical fitting on numerical results to generate accurate parameterization formulas without sacrificing details of the growth kernel. Processes considered in SNAP include fundamental aerosol processes as well as processes related to aerosol-cloud interactions. Comparison of SNAP with numerical solutions, analytical solutions, and binned aerosol model simulations showed that the new method performs well, with accuracy higher than that of the high-order numerical quadrature technique, and with much less computation time. The SNAP scheme has been implemented in regional air quality models, producing results very close to those using binned-size schemes or numerical quadrature schemes.