Gummy stem blight, caused by Didymella bryoniae (Auersw.) Rehm is a major disease of watermelon [Citrullus lanatus (Thunb.) Matsum. & Nakai] in the U.S. Plant breeders need sources of resistance that can be incorporated into adapted breeding lines to keep epidemics of this disease under acceptable control. We tested all the available accessions from the United States Department of Agriculture, Agricultural Research Service (USDA-ARS) watermelon germplasm collection, including C. lanatus var. citroides, for resistance to gummy stem blight. To perform the screen, we adopted the following protocol: 1) we used spores from a virulent isolate (or mixture of virulent isolates) grown for 2 to 3 weeks on Potato Dextrose Agar under artificial white light (12 hour photoperiod); 2) we used as inoculum a water suspension of spores in deionized water (5*105 spores/ml); 3) we ensured high relative humidity immediately after inoculation (with the presence of free-water on the leaves of test plants), by irrigating the field and using a clear plastic disease chamber in the greenhouse with artificial mist; 4) prior to inoculation we injured the trichomes of the leaves by brushing the plants with a wood stake. Our experiment was a randomized complete block with 1,332 cultigens (elite cultivars, obsolete cultivars, breeding lines, PI accessions, and cucumber checks), two assays (field and greenhouse), two or four replications, and two to six plants per plot. Cultigens were significantly more or less resistant than the resistant check PI 189225 and the susceptible check 'Charleston Gray'. PI 279461, PI 482379, PI 254744, PI 526233, PI 482276, PI 271771, PI 164248, PI 244019, PI 296332, and PI 490383 were selected as the most resistant cultigens, based on low mean rating for gummy stem blight, similar reaction in field and greenhouse, low standard deviation, and high number of replicates. The most susceptible cultigens, based on high mean rating for gummy stem blight, similar reaction in field and greenhouse, low standard deviation, and high number of replicates, were: PI 183398, PI 169286, PI 223764, PI 226445, PI 525084, and PI 534597. Further analysis of the group of the most resistant PI accessions suggested that resistance to gummy stem blight in watermelon might be determined by a single gene (or a unique set of genes) with different pleiotropic effects at the juvenile stage in the greenhouse and at the adult stage in the field. For marketing of cultivars in drier climates (i.e. southwestern U.S.), where gummy stem blight typically appears at the seedling stage during transplant production, due to high relative humidity and warm temperatures in greenhouses, it should be sufficient to develop cultivars highly resistant in the greenhouse and moderately resistant in the field. In years of extraordinary humid weather, gummy stem blight in the field would be controlled with an integrated pest management program, while in regular year of dry weather it would not affect the crop.