The commercial release of crops with engineered resistance to 2,4-D and dicamba will alter the spatial and temporal use of these herbicides. This, in turn, has elicited concerns about off-target injury to sensitive crops. In 2014 and 2015, studies were conducted in Tifton, GA, to describe how herbicide (2,4-D and dicamba), herbicide rate (1/75 and 1/250 field use), and application timing (20, 40, and 60 DAP) influence watermelon injury, vine development, yield, and the accumulation of herbicide residues in marketable fruit. In general, greater visual injury and reductions in vine growth, relative to the non-treated check, were observed when herbicide applications were made before watermelon plants had begun to flower. Although the main effects of herbicide and rate were less influential than the timing of applications with respect to plant development, the 1/75 rates were more injurious than the 1/250 rates; dicamba was more injurious than 2,4-D. In 2014, the 1/75 and 1/250 rates of each herbicide reduced marketable fruit numbers 13 to 20%, but only for the 20 DAP application. The 1/75 rate of each herbicide when applied at either 20 or 40 DAP reduced the number of fruit harvested per plot in 2015. Dicamba residues were detected in marketable fruit when the 1/75 rate in 2014 and 2015 and the 1/250 rate in 2015 was applied to plants at 40 or 60 DAP. Residues of 2,4-D were detected in 2015 when the 1/75 and 1/250 rates were applied at 60 DAP. Across both years, the maximum level of residue detected was 0.030 ppm. While early season injury may reduce watermelon yields, herbicide residue detection is more likely in marketable fruit when an off-target contact incident occurs closer to harvest.