This paper assesses the impact of aerosol properties on the formation and the development of radiation fog. Simulations were performed using the Meso-NH meteorological model including the ORILAM aerosol scheme coupled with a two-moment microphysical cloud scheme (number concentration of cloud droplets and cloud water content). The activation scheme used was taken from the work of Abdul-Razzak and Ghan (2004). "Off-line" sensitivity analysis of CCN (Cloud Condensation Nuclei) activation was performed on number, median diameter and chemical compounds of aerosols. During this "off-line" study, the interactions with the other physical processes (e.g. radiative) were not taken into account since the cooling rate was imposed. Different regimes of CCN activation and a critical value of aerosol number concentration were found. This critical aerosol number corresponds to the maximum of activated cloud droplets for a given cooling rate and given aerosol chemical properties. As long as the aerosol number concentration is below this critical value, the cloud droplet number increases when the aerosol number increases. But when the aerosol number concentration exceeds this critical value, the cloud droplet number decreases when aerosol number increases. A sensitivity study on aerosol chemical composition showed that the CCN activation was limited in the case of hydrophilic aerosol composed of material with a solubility in the 10% range. An event observed during the ParisFOG field experiment was simulated. This case took place in the polluted sub-urban area of Paris (France) characterized by particle concentrations of 17 000 aerosols per cm3. 1D simulations successfully reproduced the observed temporal evolution of the fog layer. Beyond the initial fog formation at the surface, cloud droplet formation occurred at the top of the fog layer where the cooling rate was maximum, reaching more than 10 K h 1. These simulations confirm that the aerosol particle number concentration is a key parameter for the accurate prediction of the microphysical properties of a fog layer and also influences the vertical development of fog. The important of the interaction between microphysical and radiative processes is illustrated, showing how the life cycle of a fog layer is determined by the CCN number concentration and chemical properties.