Lesion-count data on fruits/leaves from two regions of China and on leaves from controlled-environment studies were used to investigate incidence-density [incidence of leaves/fruits with lesion(s) and average number of lesions per leaf/fruit] and incidence-incidence [incidences of leaves and shoots with lesion(s)] relationships. Few of the datasets for the number of lesions per fruit/leaf could be fitted satisfactorily by a Poisson distribution. Three two-parameter distributions (negative binominal, Neyman type A and Polya-Aeppli) provided significantly better fit than the Poisson distribution, indicating a degree of aggregation in the number of lesions on a single leaf/fruit. However, many datasets could still not satisfactorily be fitted by these distributions. The dynamics of aggregation of lesions on leaves/fruits was well described by Taylor's power-law model. Regression models provided accurate predictions of the average number of lesions per leaf/fruit from the incidence of leaves or fruits with lesion(s). Nevertheless, the incidence-density relationship varied considerably between regions and between leaf and fruit scab. Field data also indicated that the number of scabbed leaves per shoot showed some degree of aggregation. The incidence of leaves with scab could be predicted accurately from the incidence of shoots with scab. The incidence-density relationships developed in this study could be used in making practical disease-management decisions when incidence of leaves with scab is less than 35%.