Morphological population balance (PB) modeling of crystallization processes predicts the dynamic evolution of shape distribution of crystals grown from solution. It requires growth kinetics of individual facets of crystals. This paper investigates model identification techniques for deriving size-dependent facet growth kinetics models as functions of supersaturation and size of the crystal. Cooling crystallizations experiments were carried out with solution concentration and temperature recorded. The seeds and obtained product crystals were characterized for shape and size distributions using an integrated particle dispersion-imaging system. The models obtained were tested using experimental data of crystallization of the same chemical under different cooling conditions. During all experiments an on-line imaging instrument was applied to obtaining real-time images for the purpose of qualitative comparison with model predicted shape distribution. (C) 2011 Elsevier Ltd. All rights reserved.