Summary. At first potyviruses were easily distinguished by biological and serological properties because only a few were known and information on their host ranges was limited. The first evidence of serological cross reaction between two of these viruses was reported in 1951 and was further corroborated for three obviously distinct members of the group in 1960. In 1968 attention was drawn to the fact that some legume and non-legume potyviruses have much wider host ranges than previously known and that within the potyvirus group there is as much biological variation within viruses and overlap between viruses as there is in serology. The concept of continuity within the group was soon supported by others and became known as the "continuum hypothesis." Results with highly sensitive serological methods using polyclonal antisera were conflicting, and nucleic acid hybridization techniques did not unambiguously discriminate between potyviruses.Recent results, obtained with antibodies directed toward epitopes located in the N -termini of the coat proteins of potyviruses, suggest that there are ways to more definitely group strains of one potyvirus and distinguish them from other potyviruses. However, there are exceptions to this rule, as in the case of bean yellow mosaic virus and clover yellow vein virus which are clearly distinct in host range, inclusion bodies, and migration velocity of coat protein, but which still react with antibodies to the N-terminal epitopes of one virus. So the question remains of whether coat-protein properties, especially the serological reactivity of N-termini, which do not alter overall virus integrity when lost, sufficiently represent the genome of a pathogenic virus entity as a single criterion for classification.