The spectral bin microphysics Hebrew University cloud model with the spatial resolution of 50 × 50 m is used to simulate the evolution of isolated deep mixed phase convective clouds under different meteorological conditions and at different aerosol concentrations. The model takes into account the effects of turbulence on droplet collision rate. Turbulent collision kernels are calculated at each time step and at each grid point. The turbulence-induced collision rate enhancement is determined by means of lookup tables calculated in the recent studies for different values of turbulent dissipation rate and the Taylor microscale Reynolds numbers. Deep convective clouds observed during the Large-Scale Biosphere-Atmosphere Experiment in Amazonia-Smoke, Aerosols, Clouds, Rainfall and Climate campaign in the Amazon region are simulated at different aerosol concentrations. Turbulence intensity in the simulated clouds is spatially inhomogeneous and reaches its maximum at the tops of multiple bubbles forming the clouds. It is shown that polluted clouds are more turbulent than those developing in the clean atmosphere. An agreement of the calculated droplet size distributions with those measured in situ is demonstrated. It is shown that turbulence accelerates formation of raindrops, especially in polluted clouds. At the same time, turbulence-induced collision enhancement lessens the amount of ice and leads to a decrease in the net accumulated rain from mixed-phase clouds. To a certain degree the effects of turbulence on precipitation counteract the aerosol effects. Since no turbulence effects on collisions of drops larger than 22 m in radius as well as on drop ice and ice collisions are considered in this study, and taking into account that a 2-D model geometry is used, the results of the study should be considered as preliminary. Additional numerical and theoretical investigations are required to quantify the turbulent effects.