Recent simulations of entangled polymer dynamics are based on slip-link models pioneered by Hua and Schieber (J. Chem. Phys. 1998, 109, 10018). In this study, we compare different slip-link models in their prediction of the linear viscoelasticity of bidisperse linear polymers to examine how effectively such models account for constraint release. We compare the DT model by Doi and Takimoto (Philos. Trans. R. Soc. London, Ser. A 2003, 361, 641), the NS model by Nair and Schieber (Macromolecules 2006, 39, 3386), and the primitive chain network (PCN) model by Masubuchi et al. (J. Chem. Phys. 2001, 115, 4387). It is found that DT and PCN models are in better quantitative agreement with literature data than is the NS model. This indicates that renewal of entanglements along the chain due to constraint release, explicitly accounted for in DT and PCN, is not equivalent to spatial fluctuations of slip-links, as included in the NS mean field approach.