Two field trials were conducted to investigate different herbage grasses and cereals for their susceptibility to the disease take-all, for their impact on concentrations of the pathogen, Gaeumannomyces graminis var. tritici (Ggt), in soil and for their effect on development of take-all in a subsequent wheat crop. In the herbage grass trial, Bromus willdenowii was highly susceptible to Ggt, produced the greatest post-senescence Ggt concentrations in soil and highest incidence of take-all in following wheat crop. Lolium perenne, Lolium multiflorum and Festuca arundinacea supported low Ggt soil concentrations and fallow the least. The relationship between susceptibility to Ggt and post-senescence concentrations in soil differed between pasture grasses and cereals. In a trial in which Ggt was added to half the plots and where wheat, barley, triticale, rye or fallow were compared, the susceptibility of the cereals to take-all was not clearly linked to post-harvest soil Ggt concentrations. In particular, triticale and rye had low and negligible take-all infection respectively, but greater postharvest soil Ggt concentrations than barley or wheat. This indicates that low Ggt concentrations on roots may build up during crop senescence on some cereals. Soil Ggt concentrations were greater following harvest in inoculated plots sown to cereals, but in the second year there was more take-all in the previously noninoculated than inoculated plots. Thus, the grass and cereal species differed in susceptibility to take-all, in their impact on Ggt multiplication and in associated take-all severity in following wheat crop.