Almost 20 years have passed since Stanley Prusiner proposed that the agent causing transmissible spongiform encephalopathies consists exclusively of a protein and termed it prion. A mixed balance can be drawn from the enormous research efforts that have gone into prion research during this time. On the negative side, the protein-only hypothesis has not been conclusively proven yet. On the positive side, our understanding of spongiform encephalopathies has experienced tremendous advances, mostly through human genetics, mouse transgenetics, and biophysical methods. Perhaps the most astonishing development is the realization that many human neurodegenerative diseases for which transmissibility has been more or less stringently excluded, may follow pathogenetic principles similar to those of prion diseases. Also, the hypothesis that prion-like phenomena may underlie certain non-genetic traits observed in yeast has resulted in the surprising recognition that the instructional self-propagating changes in protein conformation may be much more prevalent in nature than previously thought. The latter developments have been astonishingly successful, and one could now argue that the prion principle is much more solidly established in yeast than in mammals.