Levels of paspalum leaf blight (Ascochyta paspali) severity (%LAD), diseased dry matter (%DDM) and incidence on tillers and leaves on dairy pastures in northern Victoria assessed during the summer months (November-March) of 1983-84, 1984-85, 1985-86 and 1986-87 fluctuated between samplings and farms. Infected tissues accounted for 19-22% of the total dry matter in 1983-84 and 1984-85. Disease severity (%LAD) was significantly (P<0.001) linearly related to incidence of infected tillers (IT) or leaves (IL) and highest correlations were obtained following natural logarithmic transformations of severity and incidence data. Incidence of infected tillers on a 1 ha site in Northern Victoria ranged over 7- 19% in 1985-86 when samples were taken along a M-transect and 1.75-30% in 1986-87 when stratified random samples were taken using a microcomputer-based program 'Field Runner'. The latter reduced sampling bias, indicated that the distribution of disease was aggregated and a minimum sample size of 450 tillers was required. The fluctuating incidence levels with time were correlated with sampling intervals following grazing. Incidence of infected tillers on 11 farms in northern Victoria and Finley, N.S.W., in January 1987, ranged from 11.8 to 56.4%. The significance of the severity-incidence relationships, the aggregated distribution of paspalum leaf blight, variation in levels between farms, the effects of grazing on epidemics of paspalum leaf blight and crop loss and the use of 'Field Runner' in disease surveys is discussed.