In semi-crystalline thermoplastic products, final properties are strongly dependent on the thermo-mechanical history experienced by the polymer melt during processing. More precisely, structural heterogeneities such as rigidity gradients and shrinkage anisotropy are directly related to the crystalline microstructure. Therefore, accurate prediction of part properties by a processing computer simulation code requires the implementation of an appropriate crystallization kinetics model, including both the effects of thermally and flow induced structure development. One issue is the necessity to improve the modeling of shear/extensional experimental data by relating the crystallization accelerating factors to an easily accessible material related variable. Several authors modeled the effect of the flow on the crystallization kinetics by using the isokinetic approach of Nakamura. Often, the resulting kinetic equations of these models account only for the evolution of the crystallinity fraction a leaving the influence of crystalline morphology aside. We may quote the work of Guo and Narh [1], which connects the flow influence on the crystallization rate to the increase in the thermodynamic melting temperature in the Nakamura model. In 2005, R.I. Tanner presented a comparison of some models describing the polymer crystallization at low shear deformation rates under isothermal conditions. Based on Tanner's study, we developed a model of crystallization at low shearing, applied to non-isothermal flows, using only macroscopic measurable parameters. The key features of the concentrated suspension theory were used to characterize the effect of crystallization on the viscosity. In addition, we assumed that the flow generates additional crystallization nuclei via a parameter which combines the deformation and the deformation rate. The concept of germination-growth is introduced using the fundamentals of the Avrami-Kolmogorov theory, coupled with a modified Schneider's approach. The model is applied to a polypropylene, in a cooled Couette flow configuration. The results show the enhancement of the crystallization kinetics due to the shearing. The definition of global parameters simplifies the type and the number of experiments needed for the model parameter identification. The use of Schneider's approach leads to a new way of discriminating the relative roles of the flow and the temperature on the crystallization phenomenon. The competition between the two driving causes is presented and discussed: at low cooling rate or at high temperature, the shearing effect predominates. Other interesting results show the size distribution of the spherulites as well as the volume proportion for each crystalline size in the polymer. (C) 2009 Elsevier B.V. All rights reserved.