Common root rot, caused by Cochliobolus sativus, is an important disease of spring wheat and barley in the northern Great Plains. While the disease has been widely studied, extensive systematic investigation of pathological variation in C. sativus populations has not been reported. Five widely separated North Dakota counties were selected for field sampling. More than 430 isolates of C. sativus were obtained from wheat and barley plants. Pathogenicity of isolates was assessed according to amount of disease on wheat test plants in randomized, replicated greenhouse trials. Data were analyzed using the UNIVARIATE procedure of the SAS computer package. Relative pathogenicities of isolates were normally distributed continuous variables. Although the range of pathogenicity in populations of C. sativus within counties was similar, the means and variances were slightly different. Repeated planting of the same crop often results in an increase of inoculum of soilborne plant pathogens, but little is known of the effect on pathogenicity in the pathogen population. I collected C. sativus from symptomatic wheat plants from a plot planted continuously to wheat for over 90 years and from normally rotated grain fields in the surrounding area. Relative pathogenicity of isolates was tested in replicated greenhouse trials using wheat plants scored for disease six weeks after inoculation. Isolates from the continuous wheat plot were more pathogenic on average than isolates from commercial fields. Long term continuous cropping to wheat appears to shift the population of C. sativus toward more aggressive types, although many less pathogenic types also remain. Variability in C. sativus has been recognized in morphology of culture and pathogenicity for many years. Variability on virulence and/or aggressiveness of C. sativus isolates is not well documented. From the original collection of 432 isolates, 50 were selected to cover the entire range of pathogenicity. In three experiments, relative pathogenicity of the isolates was evaluated according to the amount of disease on each of four durum wheat cultivars. Aggressiveness patterns in the C. sativus-wheat system were reasonably reproducible. The C. sativus isolates originally selected for differential responses to wheat cultivars were not consistent in repeated testing. It was concluded that a range of pathogenicity (= aggressiveness ) exists in the C. sativus population and the apparent host differential responses are not repeatable and are probably experimental artifacts.