Watermelon [Citrullus lanatus (Thunb.) Matsum. & Nakai] is a major vegetable crop in the United States. Powdery mildew [ Podosphaera xanthii (Castagne) Braun & Shishkoff (syn. Sphaerotheca fuliginea auct. p.p.)] is now a common disease of watermelon in the United States and other parts of the world. The disease is of great economic importance because it leads to loss in fruit yield and quality. The watermelon germplasm collection has not been screened for resistance to powdery mildew race 2W. It would be useful to obtain a source of resistance to powdery mildew for incorporation into commercial cultivars. The objectives of this study were (1) to develop an efficient and reliable method for screening for powdery mildew resistance; (2) to screen the USDA watermelon germplasm collection along with available cultivars for resistance to powdery mildew race 2W; (3) to determine the heritability and genetic variance estimates for resistance to powdery mildew race 2W. Two experiments were carried out for development of a method for screening for resistance to powdery mildew. In these experiments, effect of humidity, inoculum source and application method, and type of growth chamber on disease severity were tested. Consistent and significant differences in disease severity among four resistant and susceptible accessions were observed with the use of a spray of spore suspension of inoculum and when plants were maintained at normal greenhouse relative humidity. No supplementation of the relative humidity was required. The use of a special humidity chamber and maintenance of spores on either squash or watermelon plants did not have significant effects on disease severity. In the germplasm screening experiment the entire available U.S. Plant Introduction collection of Citrullus Schrad. ex Eckl. & Zeyh. species was screened. Resistance was found in wild type accessions and was characterized by moderate to high variability. Eight of the 1,654 accessions and cultivars demonstrated high levels of resistance with low levels of phenotypic variability. Using data from the screening and retest studies, the most resistant accessions were PI 632755, PI 386015, PI 346082, PI 525082, PI 432337, PI 386024, and PI 269365 and PI 189225. Twenty-three accessions demonstrated uniform intermediate resistance. Inheritance of powdery mildew race 2W was studied in two accessions, PI 189225 and PI 270545. Two susceptible parent lines PI 269677 and ‘Charleston Gray’ were crossed with the resistant accessions. Parents, F₁, F₂ and backcross populations were evaluated in the greenhouse. Inheritance of resistance in PI 189225 was found to be multigenic, while in PI 270545, a major gene was found whose expression was more complex than that suggested by the single gene alone. No discrete phenotypic classes were observed in the segregating F2 populations of any cross. Based on quantitative analyses, each resistant line contained at least two to three effective factors for powdery mildew resistance. Additive gene action was the most important component of variation in all three crosses. Dominance effects were not significant. Narrow sense heritability estimates for powdery mildew resistance was 0.62 for PI 189225 and 0.92 for PI 270545.