Field experiments were carried out to measure the effect on the yield and composition of sugar beet of infection with beet yellows virus, and to determine how the effect varies with rate and date of infection and with date of sowing and harvesting of the crop.

Control measures designed to prevent spread of infection within experimentally infected plots and introduction of infection from sources outside the experiment, were not completely effective. Consequently the numbers of plants which became infected differed from those prescribed by the experimental treatments. The effects of infection were, therefore, estimated by means of regressions on the numbers of plants observed to become infected on successive dates.

After correcting the observed yields for accidental infections by means of the regression coefficients, it was found that the effect of infection on the yield of sugar was linearly related to the rate of infection, i.e. the loss of yield of sugar caused by infection was proportional to the percentage of infected plants. There was no compensation for loss in infected plants by increased growth of healthy neighbours.

The effect of infection on yield of sugar decreased linearly with later date of infection, falling to values not significantly different from zero at the date of harvest. The loss of yield caused by infection was therefore approximately proportional to the interval of time between the date of infection, as shown by the appearance of symptoms, and the date of harvest.