Fiji disease virus is a propagative, persistently transmitted virus that multiplies in species of the delphacid planthopper genus Perkinsiella, and in sugarcane, the feeding host of the insect. Efforts to improve and modify the disease rating system for Fiji disease have largely focussed on the planthopper as individual vectors of the virus, rather than as a population of the principal, or at least an alternative, host of the virus. This perspective has resulted in key parameters of disease incidence resulting from plant infection by propagative, persistently transmitted viruses being largely overlooked or misunderstood during efforts to improve the rating system. These parameters include the relatively long acquisition, latency, and transmission times, the percentage of the population containing virus, or viruliferous, in the above periods, and the effects of population density and number of plants visited on disease incidence. Suggestions to modify trial design to improve virus transmission to the plant, based on the disease incidence parameters of the propagative, persistent transmission class, are presented and the practical difficulties of implementing these proposals are discussed. In the context of fully understanding the underlying biology of this virus–insect–plant system, the hypothesis that Fiji disease virus, as a plant-infecting member of the Reoviridae, is primarily an insect virus with a secondary plant host, and may have diverged from an insect-infecting virus relatively recently is proposed and compared with other members of the family Reoviridae.