An analysis of new particle formation probability in the marine boundary layer (MBL) is conducted using a detailed aerosol dynamics and gas-phase chemistry model, thermodynamically correct classical binary (H2O-H2SO4) nucleation theory, and recently developed tel nary (H2O-H2SO4-NH3) nucleation theory. Additionally, the effect of boundary-layer meteorology (i.e., adiabatic cooling, small scale fluctuations, and entrainment) in enhancing nucleation is also examined. The results indicate that for typical marine conditions, binary nucleation does not occur for any realistic conditions regardless of adiabatic cooling, turbulent fluctuations, or entrainment. For polar marine conditions, binary nucleation does occur due to lower temperatures, and is enhanced due to turbulent fluctuations. An increase in detectable particle sizes (N-3>3 nm), is only seen after multiple boundary layer circulations for conditions of high dimethyl sulphide (DMS) concentrations (400 ppt). Under extreme conditions of entrainment of fi ee-troposphere layers containing very low aerosol condensation sinks and extraordinary high sulphuric acid concentrations (>10(8) molecules cm(-3)), increases in detectable particles up to 10,000 cm are predicted only in polar marine air, but are viewed as unlikely to occur in reality. Comparison of model simulations with observed values of DMS and sulphuric acid in polar marine air masses suggest that binary nucleation may lead to an enhancement of approximate to 1000 cm(-3) in N-3 particle concentration, but not to enhancements of approximate to 10,000 cm(-3) Ternary nucleation is predicted to occur under realistic sulphuric acid (1.2 x 10(7) molecules cm(-3)) and ammonia (>5 ppt) concentrations; however, significant growth to detectable sizes (N-3) only occurs for DMS concentrations of the order of 400 ppt and very low aerosol condensation sinks, but these conditions are thought to be very infrequent in the MBL and are unlikely to make a significant contribution to the general MBL aerosol concentration. It is plausible that the background MBL aerosol concentration could be maintained by a slow, almost undetectable production rate, and not by noticeable nucleation events where large enhancements in N-3 concentrations are observed. The former requires sustained DMS concentrations of the order of 100 ppt which seems unlikely. In summary, the occurrence of new particles in the unperturbed MBL would be difficult to explain by DMS emissions alone. DMS emissions can explain the occurrence of thermodynamically stable sulphate clusters, but under most conditions, to grow these clusters to detectable sizes before they are scavenged by coagulation, an additional condensable species other than DMS-derived sulphuric acid would be required. In the event, however, of significant removal of the preexisting aerosol due to precipitation, the MBL aerosol can be replenished through growth of new particle formed through ternary nucleation under moderately high DMS concentrations.