A three-dimensional mesoscale meteorological model was used to study the interplay between the dynamical (turbulent mixing and advection) and physico-chemical (sea-air flux and photochemical sink by OH) processes that control dimethylsulfide DMS concentrations and their distribution in the marine boundary layer (MBL) during the First Aerosol Characterization Experiment ACE 1. Atmospheric DMS concentrations were constrained using observed seawater DMS concentrations and box model derived OH concentrations. Lateral boundary values of dynamical parameters were derived from the 6-hourly meteorological analysis of the European Centre for Medium-Range Weather Forecasts. Calculated DMS concentrations, wind speed and direction, and cloud cover were compared with measurements made aboard the R/V Discoverer and on the three NCAR/C130 aircraft flights during the LagB experiment. Model-generated atmospheric DMS concentrations agreed with the DMS observations from the NCAR/C130 aircraft flights during the LagB experiment (R2=0.69) assuming OH is the only oxidant and DMS flux parameterization based on Liss and Merlivat <1986>. Comparison with Eulerian measurements made aboard the R/V Discoverer showed that the model simulated the range of observed values but not the hour-to-hour variation observed in the atmospheric DMS concentrations. Part of the discrepancy was attributed to uncertainties in DMS sea-to-air transfer velocity, small scale features of seawater DMS that are beyond the model resolution, and uncertainties in the venting of the boundary layer by shallow clouds. A quantitative budget at the ship location revealed a strong impact of advection processes in determining DMS levels and temporal evolution. The three-dimensional mesoscale meteorological model was also used to estimate the effect of the low spatial resolution used in global models on seawater DMS concentrations and atmospheric OH concentrations. ¿ 1998 American Geophysical Union