The microstructural, mechanical, compositional, and degradative properties of a nerve conduit are known to strongly affect the regenerative process of the injured peripheral nerve. Starting from the fabrication of micropatterned collagen-based nerve guides, according to a spin-casting process reported in the literature, this study further investigates the possibility to modulate the degradation rate of the scaffolds over a wide time frame, in an attempt to match different rates of nerve regeneration that might be encountered in vivo. To this aim, three different crosslinking methods, that is, dehydrothermal (DHT), carbodiimide-based (EDAC), and glutaraldehyde-based (GTA) crosslinking, were selected. The elastically effective degree of crosslinking, attained by each method and evaluated according to the classical rubber elasticity theory, was found to significantly tune the in vitro half-life (t1/2 ) of the matrices, with an exponential dependence of the latter on the crosslink density. The high crosslinking efficacy of EDAC and GTA treatments, respectively threefold and fourfold when compared to the one attained by DHT, led to a sharp increase of the corresponding in vitro half-lives (ca., 10, 172, and 690 h, for DHT, EDAC, and GTA treated matrices, respectively). As shown by cell viability assays, the cytocompatibility of both DHT and EDAC treatments, as opposed to the toxicity of GTA, suggests that such methods are suitable to crosslink collagen-based scaffolds conceived for clinical use. In particular, nerve guides with expected high residence times in vivo might be produced by finely controlling the biocompatible reaction(s) adopted for crosslinking.