There are two central phenomena in prion disease: prion replication and prion neurotoxicity. Underlying them both is the conversion of a host-encoded ubiquitously expressed protein, prion protein (PrP(C)), into a partially-protease resistant isoform, PrP(Sc), which accumulates in the brain. PrP(Sc) is associated with both pathology and infectivity. In the absence of PrP(C), PrP(Sc) cannot be generated and PrP-null mice do not propagate infectivity or develop pathology on infection with scrapie. However, while PrP(C) expression is fundamental to both prion infectivity and neurodegeneration, the uncoupling of these processes is a growing concept in the field. This dissociation is evident in subclinical states of prion infection, where PrP(Sc) levels are high in the absence of disease, and in transgenic mice experiments, where, despite extra-neuronal PrP(Sc) accumulation, even in very high amounts, there is no neurotoxicity. Both these models have further implications. Thus depleting PrP(C) from neurons (but not glia) of prion-infected mice prevents clinical disease, and detaching it from the surface of cells by removing its anchor does the same. The uncoupling toxicity and infectivity has implications for the nature of the neurotoxic species; these mouse models suggest that the site for the generation of this species is intra-neuronal. This review considers the role of neuronal surface-expressed PrP(C) in mediating toxicity in prion infection, and the dissociation of this from the deposition of PrP(Sc).