Studies of spatiotemporal dynamics are central to efforts to characterize the epidemiology of infectious disease, such as mechanism of pathogen spread and pathogen or vector sources in the landscape, and are critical to the development of effective disease management programs. To that end, we conducted a multi-year study of 20 vineyard blocks in coastal northern California to relate the dynamics of a mealybug vector, Pseudococcus maritimus (Ehrhorn) (Hemiptera: Pseudococcidae), to incidence of grapevine leafroll disease (GLD). In each vineyard block, a subset of vines were scored visually for relative mealybug abundance, disease was quantified by visual assessment, and virus presence was verified using standard laboratory molecular assays. GLD incidence was analyzed with a classification and regression tree, and with a hierarchical model that also captured variability among blocks and heterogeneity within blocks. Both analyses found strong interannual variability in incidence, with the hierarchical model also capturing substantial between- and within-block heterogeneity, but with significant contributions of vector abundance and pathogen supply (prior disease incidence) to the frequency of newly diseased vines. These results strengthen further the conclusion that mealybug vectors are causally related to pathogen spread in this system and are therefore an important target for management. Moreover, they are consistent with relatively efficient secondary spread of the pathogen, suggesting an important role for the removal of diseased vines as a tool to mitigate further damage.