Aquatic birnaviruses infect a large variety of fish, molluscs and crustacea in the freshwater, estuarine and marine environments throughout the world and yet despite this extensive host range and geographical distribution, the vast majority of isolates have been found to be antigenically-related to the original reference serotypes (VR299, Sp and Ab) of infectious pancreatic necrosis virus (IPNV) of salmonids. However, in early studies it was seen that there is a high degree of antigenic diversity amongst isolates with some reacting only relatively weakly with these three traditional serotypes, suggesting that other serotypes may exist. In this review we examine the published studies on the degrees of antigenic relatedness between IPNV isolates and other aquatic birnaviruses and discuss the attempts by various authors to identify and/or define different serotypes. The published data from serum neutralization tests is critically appraised and the factors influencing neutralization results discussed. Details are then presented of our own cross-neutralization study with almost 200 isolates of IPNV and other aquatic birnaviruses using a standardized procedure. On the basis of the limit we set for the maximum degree of difference from a reference serotype for an isolate to belong to that serotype, the isolates we examined could be divided into two distinct serogroups (which do not cross-react), the first containing nine serotypes and the second ontaining a single serotype. We propose a change from the traditional terminology of the recognised referenreference serotypes offers tsubstantiated the serotype classification scheme achieved by serum neutralization tests.