Amphotericin B (AmB) has been shown to delay hamster scrapie. Infectivity studies have been performed previously using AmB in order to understand the relationship between the accumulation of an abnormal isoform (PrPres) of the prion protein and 263K scrapie agent replication in the brain. The first study reported that AmB had no effect upon agent replication, although it delayed the development of both clinical signs and PrPres accumulation. However, subsequent experiments using the same model showed a significant effect both on agent replication and PrPres accumulation early in infection. This fundamental discrepancy was assumed to be linked to differences in experimental protocols. In order to unravel the issue, a new experiment has been performed encompassing different protocols and using an AmB derivative, MS-8209, that can be used at higher doses because of its lower toxicity. The findings of this study exclude the suspected differences in the protocols as the reason for previous conflicting results, and suggest strongly that these discrepancies were due to a low dose of AmB causing a 'threshold effect'. Overall, this study indicates that, in this model, PrPres cannot be dissociated from infectivity by polyene antibiotics.