In order to exploit the information contained in the particle size distributions obtained from batch precipitation experiments and derive reliable kinetic parameters estimates, a method based on reaction modelling and parameter optimization has been developed. It has been shown that the experimental reator set-up should be modelled as a plug flow reactor followed by a fed-batch cell. The population balance equation describing this system is discretized by using a method of classes to generate a system of ordinary differential equations and the latter is solved by an adaptive step Runge-Kutta method. On top of this, a direction set optimization method, coupled to the intelligent use of some physical insight into the system, is used to iterate on the kinetic parameter values until the computed particle size distribution best fits the experimental one.