Atmospheric production of cloud condensation nuclei (CCN) becomes increasingly important when going to rural or more remote locations. A preliminary investigation on the relative roles of two of the most important compound classes contributing to atmospheric CCN production, sulfate and secondary organic matter, was performed. The investigation was made using a Lagrangian air parcel model that simulates the time evolution of a particle population within a marine or a continental boundary layer. Model simulations demonstrated that for a system containing small (<40--50 nm diameter) primary or secondary particles, production of new CCN can be thought of as a two-step process: (1) condensational growth of very small particles to a so-called potential activation zone, in which these particles can be activated to cloud droplets at high supersaturations associated with strong updrafts, and (2) growth of particles from the potential activation zone to sizes where they form droplets at more typical cloud supersaturations. Since the exact location and width of the potential activation zone depend on both cloud physics and the chemical properties of the smallest particles able to act as CCN, the border between these two steps is not very sharp under real atmospheric conditions. According to the simulations, the first step of the above steps relies heavily on the presence of nonvolatile organic vapors, with their required gas phase source rate being of the order 0.1--0.5 μ g m-3 d-1 or greater. Sulfuric acid formed in the gas phase can give significant contribution under very specific conditions only. While condensational growth associated with secondary organics may also be involved in the second step, in-cloud sulfate production associated with strong updrafts appears to be the most effective way to move particles from the potential activation zone to larger sizes. In general, the second step was shown to depend on the amount of hygroscopic material transported to particles in the potential activation zone, on the variability of the cloud supersaturation between successive cloud events and on the amount of SO2 available for in-cloud oxidation. ¿ 2001 American Geophysical Union