`TifEagle' (2n = 3x = 27) hybrid bermudagrass [Cynodon dactylon (L.) Pers. (2n = 4x = 36) × Cynodon transvaalensis Burtt-Davy (2n = 2x = 18)] is an ultradwarf cultivar for greens, and `TifSport' (2n = 3x = 27) is a more versatile hybrid used on fairways, athletic fields, and lawns. To develope a transformation system and determine if somaclonal variation was present in regenerated plants, both cultivars were tested for their ability to produce embryogenic callus from which plants could be regenerated. Sliced nodes of both cultivars and immature inflorescences from `TifSport' were used as the explant sources. Cultures were initiated on Murashige and Skoog medium supplemented with 6.79 μm 2,4-D and 0.044 μm BA (`TifSport' and `TifEagle') or 6.79 μm 2,4-D plus 200 mg.L-1 casein hydrolysate (`TifSport'). In total, 51 plants were regenerated from callus of a single node of `TifEagle'. Nodes from `TifSport' did not produce embryogenic callus. In total, 29 plants were regenerated from callus of `TifSport' produced from immature inflorescences. These plants were grown in the field for at least one season, and 5-cm-diameter plugs were harvested, repotted in a greenhouse, and allowed to reestablish. Data on canopy height, leaf width, leaf length, and number of stolons were collected. Seven `TifEagle'-derived entries (14%) were not significantly different (α = 0.05) from `TifEagle' harvested from the breeder plot in Tifton, Ga., for all measured traits, and 41%, 24%, and 22% differed by one, two, or three measurements, respectively. Flow cytometry indicated that 33% (13 plants) of the `TifEagle' regenerants were hexaploid (2n = 6x = 54) and the rest remained triploid. One `TifSport' regenerant was significantly different (α = 0.05) for plant height.