There have been many advances in testing procedures to detect seed-borne virus infection in seed samples. However, scant attention has been given to the implications of improved test results in terms of the economic losses resulting from sowing seed stocks with different amounts of infection. For agricultural and horticultural industries to use the results of tests on representative samples, defined 'threshold' values for percentage seed infection are required that identify acceptable levels of risk of economic losses resulting from sowing the virus-infected seed stocks. Such information is provided by field experiments in which infected seed is sown and the consequences are followed in terms of virus spread, yield losses and infection of newly produced seed. These field experiments need to continue over several years at diverse sites so that they represent a wide range of infection scenarios. Extensive surveys to determine seed-borne virus occurrence in different regions are also required to define areas of greater or lesser risk of economic losses. In this paper, an example is described of how field experiments and surveys were used to define 'threshold' values of seed-borne Cucumber mosaic virus infection in an annual crop (lupin: Lupinus angustifolius) and two such examples are given for pasture species: Cucumber mosaic virus in subterranean clover (Trifolium subterraneum), and Alfalfa mosaic virus in annual burr medic (Medicago polymorpha). The aim of this paper is to encourage others to address the urgent need for similar 'threshold' information with other economically important combinations of seed-borne viruses and host plant.