Cucurbit downy mildew (CDM), caused by the obligate oomycete Pseudoperonospora cubensis, has resurged around the world during the past three decades. A new pathotype or genetic recombinant of P. cubensis have been suggested as possible reasons for the resurgence of CDM in the United States in 2004. In total, 22 isolates collected between 2004 and 2014, mainly in the eastern United States, were tested for their compatibility with a set of 15 cucurbit host types. The virulence structure within these isolates was evaluated on a set of 12 differential genotypes from eight genera. All isolates were highly compatible with the susceptible cultivar of Cucumis sativus, whereas the least compatibility was observed with Luffa cylindrica and Momordica charantia. Based on the compatibility with the differential host set, five pathotypes (1, 3, 4, 5, and 6) were identified among the 22 isolates examined. Pathotypes 1 and 3 had not been previously described in the United States and isolates of these two new pathotypes were also compatible with 'Poinsett 76', a cultivar of C. sativus known to be resistant to CDM prior to 2004. Virulence within the pathogen population was expressed based on virulence factors, virulence phenotypes, and virulence complexity. The number of virulence factors ranged from two to eight, indicating a complex virulence structure, with 77% of the isolates having five to eight virulence factors. Thirteen virulence phenotypes were identified; the mean number of virulence factors per isolate and mean number of virulence factors per virulence phenotype was 5.05 and 5.23, respectively, indicating that complex isolates and phenotypes contributed equally to the complex virulence structure of P. cubensis. Gleason and Shannon indices of diversity were 3.88 and 2.32, respectively, indicating a diverse virulence structure of P. cubensis within the United States population. The diverse virulence and high virulence complexity within the pathogen population indicate that host resistance alone in available cucurbit cultivars will not be effective to control CDM. An integrated approach involving a combination of fungicide application and introduction of cultivars with new resistance genes will be required for effective management of CDM.