The discovery of prion disease transmission in mammals, as well as a non-Mendelian type of inheritance in yeast, has led to the establishment of a new concept in biology, the prion hypothesis. The prion hypothesis postulates that an abnormal protein conformation propagates itself in an autocatalytic manner via recruitment of the normal isoform of the same protein as a substrate, and thereby acts either as a transmissible agent of disease (in mammals) or as a heritable determinant of phenotype (in yeast and fungus). Although reconstitution of fully infectious PrP^Scin vitro from synthetic components has not yet been achieved, numerous lines of evidence indicate that the prion protein is the major and essential component, if not the only one, of the prion infectious agent. This article summarizes our current knowledge about the chemical nature of the prion infectious agent, describes potential strategies and challenges related to the generation of prion infectivity de novo, proposes new hypotheses to explain the apparently low infectivity observed in the first synthetic mammalian prions, and describes plausible effects of chemical modifications on prion conversion.