Crop rotation is the oldest, and perhaps the best cultural practice for reducing the risk of take-all. The effects of crops sown before wheat in a rotation are known in detail, but we know little about the opportunities for reducing take-all risk by planting certain crops in the summer period between wheat harvest and the planting of a subsequent winter wheat crop. We investigated the effects on take-all of five summer fallow crops, two soil tillage treatments and a fungicide seed treatment, in a five site-year experiment. We tested the effects of oats, oilseed rape, mustard, ryegrass and volunteer wheat crops. Bare-soil plots were also included. Take-all epidemics varied with year and site. Summer fallow crops had a greater effect on tilled plots. The incidence and severity of take-all were significantly higher in the wheat volunteer plots, whereas maintaining bare soil provided the lowest level of disease. Oilseed rape had no significant effect on take-all incidence in our experiment. The best candidates for reducing take-all risk appeared to be oats, mustard and ryegrass. These summer fallow crops decreased disease levels only when associated with conventional tillage. Summer fallow crops did not alter take-all decline in the same way as a break crop after a wheat monoculture.Abbreviations: AUDPC -area under the disease progress curve; Ggt -Gaeumannomyces graminis var. tritici; GLM -generalised linear model; Gri201, Gri601, Gri702, LR202 and LR203 -field experiment codes (Gri for Grignon and LR for Le Rheu, followed by the order of the assessed wheat crop [2 for second wheat crop, 6 for sixth wheat crop or 7 for seventh wheat crop] and the harvest year [2001, 2002, or 2003]); GS -Zadoks growth stage; GS 15 -5 leaves unfolded; GS 30 -pseudostem elongation; GS 32-33 -midstem elongation; GS 65 -flowering; GS 89 -ripening; F+/F) -with/without fungicide seed treatment; Till/no-till -conventional tillage/direct drilling.