This study presents continued development of the Nenes and Seinfeld (2003) cloud droplet activation parameterization. First, we expanded the formulation to (1) allow for a lognormal representation of aerosol size distribution, and (2) include a size-dependant mass transfer coefficient for the growth of water droplets to accommodate the effect of size (and potentially organic films) on the droplet growth rate. The performance of the new scheme is evaluated by comparing the parameterized cloud droplet number concentration with that of a detailed numerical activation cloud parcel model. The resulting modified parameterization robustly and closely tracks the parcel model simulations, even for low values of the accommodation coefficient (average error 4.1 ± 1.3%). The modifications to include the effect of accommodation coefficient do not increase the computational cost but substantially improve the parameterization performance. This work offers a robust, computationally efficient and first-principles approach for directly linking complex chemical effects (e.g., surface tension depression, changes in water vapor accommodation, solute contribution from partial solubility) on aerosol activation within a global climate modeling framework. Copyright 2005 by the American Geophysical Union.