In 1975, 1976 and 1977 counts were made of cereal aphid populations in winter wheat fields in West Sussex and Norfolk. Data were also obtained on the area of cereals treated with insecticides to control aphids in different years, dates of application and the types of insecticide used. Most crops were treated when aphid populations were at their peak or had already started to decline. A model which simulates the damage done by Sitobion avenae (F.) to winter wheat was used to analyse the survey data and revealed a large difference between potential and achieved profits. It also showed how the value of insecticide treatment varies throughout the course of an infestation and how the value of a forecasting system depends on its timing and accuracy, the size of the aphid outbreak and its probability of occurrence.