An explicit representation of the liquid-water-drop size Spectrum is coupled to two- and three-dimensional boundary layer models for the purposes of studying the dynamics of precipitating stratocumulus. A number of hypotheses regarding the effect that precipitation has on boundary layer structure are explored. The use of the three dimensional model to do large-eddy simulation (LES) of the stratocumulus topped boundary layer suggests that the hypothesis that precipitation unambiguously leads to shallower c2ouds and shallower boundary layers is incorrect. Comparisons of LES with and without drizzle also indicate that the tendency of drizzle to generate internal stratification within the boundary layer need not result in the bo·1ndary layer becoming decoupled. Fluxes may actually become stronger, although more intermittent, in response to the moistening and cooling of the sub-cloud layer relative to both the cloud and surface-- although precipitation forming in clouds initially un-coupled to the surface may lead to solutions with a different character. Overall we find that the primary consequence of drizzle resides in its tendency to stabilize the sub-cloud layer with respect to the cloud layer. The dynamical consequences of this are the reduction in the buoyancy flux , smaller values of vertical velocity variance but larger skewness, and a decline in entrainment rates. Consequently, precipitation affects cloud structure indirectly by reducing cloud top entrainment rates. Drizzle also directly modifies cloud structure by partitioning the boundary layer into a cooler-moister sub-cloud layer and a warmer-drier cloud-layer. This leads to reduced values of cloud liquid water relative to what is expected for a well mixed layer. Depending on the nature of the cloudtop thermodynamic jumps the indirect and direct effects of drizzle on cloud structure, may either cooperate or compete. Changes in boundary layer structure due to precipitation tend to have a forced character which approaches an equilibrium balance between turbulent transport and drizzle on the eddy turn-over time-scale. The simulated boundary layer relaxes back to a state characteristic of non-precipitating boundary layers on a time-scale of a few hours after the cessation of precipitation . A number of hypotheses are formed and discussed on the basis of two dimensional integrations, one of which is that the more cumulus-like dynamics characteristic of strongly precipitating boundary layers favors up-scale growth in the characteristic horizontal size of convecting elements. Such up-scale growth is not seen in non-precipitating solutions. Also the dynamics of mixed layer models are analyzed and found to predict the wrong sensitivity of surface fluxes to precipitation. A new two layer model is proposed which attempts to correct this short-coming.