Coalescence in polymer blends was modeled by combining population dynamics with three coalescence theories: Smoluchowski, trajectory, and film drainage. Modeling results are compared with our previous coalescence experiments on polystyrene/high density polyethylene blends. Comparison of the Smoluchowski theory (no hydrodynamic interactions) with our experiments reveals that ideal collisions are the main mechanism of coalescence. Coalescence efficiency, however, decreases with increasing shear rate and particle-size difference. Most interestingly, increasing the ratio of particle to matrix viscosity causes coalescence efficiency first to increase and then to decrease. The effects of particle-size difference on coalescence efficiency can be modeled by considering changes in trajectory due to hydrodynamic interactions between particles. By considering particle deformation, film drainage theory predicts that coalescence efficiency decreases with increasing shear rate. Comparison with the experiments was only qualitatively correct. However, the prediction that coalescence efficiency decreases with increasing the viscosity ratio by both the trajectory theory and film drainage theory is not consistent with the experimental results. The reasons for the disagreements are discussed in terms of large deformation of colliding particles.