Precipitation in stirred vessels is a complex process involving interactions of physico-chemical nucleation and particle growth with fluid mixing. The simulation of precipitation via full CFD analysis still presents challenging computational difficulties. This paper seeks to examine how the fluid mixing influences the crystal size distribution (CSD) by applying simplified fluid mixing models. These can quantify how the separate effects of overall fluid convection, degrees of fluid macro-mixing and the extent of feed-separation determine the CSD. A new approach for describing particle growth has been developed. This model can solve the dynamic population balance equations by integration of ordinary differential equations. The model ensures the number balance as well as a mass balance on the solid, using a free discretised length and approach. Then, the simulation under both a set of representative fluid mixing cases which varies the extent of macro-mixing (by sets of compartments or zones) and the different degrees of feed-separation (by different addition modes of the two reagents) were performed. It is predicted that joint feeds will give a larger and wider CSD, on the other hand, the most 'extremely' separated feeding positions can compensate for the weakness of the macro-mixing. From this practioners will be able to choose the balance of stirring intensity and feed policies to increase or reduce crystal size, as well as adjust the shape of the CSD itself for double-jet operation. (c) 2005 Elsevier Ltd. All rights reserved.