Sexed sperm in dogs is of interest because of being polytocous, and as a result, the greatest number of offspring of the same sex can improve the market, although few studies assessing sperm sexing have been performed in this species. The present study, therefore, was conducted to evaluate the effects on sperm quality and the effectiveness of three discontinuous density gradients to separate dog sperm containing X and Y chromosomes. Thirty ejaculates from ten adult dogs were collected by digital manipulation of the penis. Cells were separated using gradients of Percoll® and Percoll® associated with Nycodenz® or Ficoll. The cells were evaluated for motility by the CASA system (Computer-Aided Semen Analyzer) and for concentration and recovered sperm concentration (after centrifugation), sperm morphology, plasma and acrosomal membrane integrity, and mitochondrial function pre- and post-centrifugation. The percentage of sperm containing X and Y chromosomes was also evaluated pre- and post-centrifugation by quantitative real-time PCR (qPCR). The use of the Ficoll gradient resulted in the greatest sperm quality after centrifugation; however, no sperm enhancement containing X or Y chromosome occurred with use of any of the methods (Percoll® 54.8 ± 1.9 compared with 45.2 ± 1.9; Percoll® associated with Nycodenz® 53.2 ± 2.0 compared with 46.8 ± 2.0; and Percoll® associated with Ficoll 55.0 ± 1.5 compared with 45.0 ± 1.5 for the percentages of cells containing the X and Y chromosomes, respectively). Thus, it was concluded that the technique of sexing dog sperm using density gradients was not effective for commercial application.