During the summer of 1987 a survey of dieback in non-woodland ash trees was undertaken in Great Britain. After excluding certain areas due to their known low ash population, two hundred 10 km squares were visited and detailed data collected on the condition of ash in a plot selected within each square. Information was obtained on 4454 ash trees, and also on 1022 oak trees which were encountered in the plots. The overall incidence of dieback was 19% in the sample of ash and 18% in the sample of oak. In ash, the condition was found to occur mainly in the east of the country, with areas of greatest damage, both in terms of proportion of trees affected and degree of crown loss, being found in the south-east Midlands. The healthiest trees were found in the west of the country, in particular in Wales. Large ash trees were found to be suffering more from dieback than small trees and single trees more than trees in groups. The condition was found to be more common on rendzina and gley soils than on brown earths, podzols and ‘unclassified’ soils. Trees in the countryside had a much higher incidence of dieback than did urban trees (20% compared with 11% respectively). Among rural trees, associations were found with various kinds of current agricultural practice. Thus the incidence of damage in trees surrounded by arable land was 38% compared with 10% for trees surrounded by grassland. There was also a greater incidence of damage in trees near to roads (23%) than in trees away from roads (12%). Finally, where there was an adjacent ditch, the likelihood of dieback was substantially increased. The data for oak followed a similar pattern to ash except that no effect of the presence of a ditch could be detected. Correlations of damage with rainfall and pollution variables were calculated. Ash dieback was negatively correlated with rainfall and several relationships emerged with pollution variables; the strongest being positive relationships for ‘arable only’ trees. With oak some negative correlations were found with rainfall and pollution.