The dynamics of aerosols in the marine boundary layer (MBL) are simulated with the marine boundary layer aerosol model (MARBLES), a one-dimensional, multicomponent sectional aerosol model <Fitzgerald et al., this issue; Gelbard et al., this issue>. First, to illustrate how the various aerosol processes influence the particle size distribution, the model was run with one or two processes operating on the same initial size distribution. Because of current interest in the effects of cloud processing of aerosols and exchange of aerosols with the free troposphere (FT) on marine aerosol size distributions, these two processes are examined in considerable detail. The simulations show that the effect of cloud processing (characteristic double-peaked size distribution) in the upper part of the MBL is manifested at the surface on a timescale that is much faster than changes due to exchange with the FT, assuming a typical exchange velocity of 0.6 cm s-1. The model predicts that the FT can be a significant source of particles for the MBL in the size range of the cloud-processing minimum, between the unactivated interstitial particles and the cloud condensation nuclei (CCN) which have grown as a result of conversion of dissolved SO2 to sulfate in cloud droplets. The model was also used to simulate the evolution of the aerosol size distribution in an air mass advecting from the east coast of the United States out over the ocean for up to 10 days. The modification of a continental aerosol size distribution to one that is remote marine in character occurs on a timescale of 6--8 days. Nucleation was not observed in the base case 10-day advection simulation which assumed rather typical meteorological conditions. However, significant nucleation was predicted under a more favorable (albeit, atypical) combination of conditions which included significant precipitation scavenging (5 mm h-1 of rain for 12 hours), colder temperatures by 10 ¿C (283 K at the surface decreasing to 278 K at 1000 m) and a high DMS flux (40 μmol m-2 d-1). In a test of model self initialization, long-term (8--10 days) predictions of marine aerosol size distributions were found to be essentially independent of initial conditions. ¿ 1998 American Geophysical Union