Low-level incorporation of impurities into industrial crystallizer products can lead to several undesirable properties such as high product colour, low bulk density and poor flow properties. The driving force for impurity incorporation into a crystal may be either thermodynamic or kinetic or a blend of both. Models for each mechanism are reviewed and tested on a set of experimental data. The model system chosen to generate data with which to compare incorporation models is a galactitol/water system with lower concentrations of other alditols (mannitol, xylitol, sorbitol) present at concentrations below their saturated concentrations. Galactitol was crystallized from solution by primary nucleation and the batch was subsequently cooled to increase yield of galactitol. Analytical results showed that all of the lower concentration impurities had co-crystallized to some degree with the galactitol substrate. The incorporation rate was found to be greatest during the primary nucleation event where kinetic considerations dominated and lower during the cooldown phase where thermodynamic considerations dominated. A thermodynamic model was found best able to model the overall incorporation behaviour with empirical estimation of parameters. A Langmuir approach was found to be able to model the data satisfactorily, but when extended to model the incorporation behaviour in an independent dataset, it did not perform as well as the thermodynamic model.