The effects of take-all epidemics on winter wheat yield formation were dete
rmined, and disease-yield relationships were established to assess the agro
nomic efficacy and economic benefits of control methods. Epidemics were gen
erated in naturally infested fields by varying cropping season, crop order
in the rotation, and experimental fungicide seed treatment. Disease inciden
ce and severity were assessed from tillering to flowering. Yield components
were measured at harvest. Models simulating the formation of the yield com
ponents in the absence of limiting factors were used to estimate the losses
caused by take-all. Losses were predicted by the disease level at a specif
ic time or the area under the disease progress curve, reflecting accumulati
on during a specific period. Losses of grain number per square meter and 1,
000-grain weight were linked to cumulative disease incidence between the be
ginning of stem elongation and flowering, and disease incidence at midstem
elongation, respectively. Yield losses were accounted for by both cumulativ
e disease incidence between sowing and flowering, and disease incidence at
midstem elongation. Results confirm the importance of nitrogen fertilizatio
n in reducing the impact of take-all on wheat.