Misassignment between DNA samples and clinical or epidemiological data may compromise the results of genetic association studies. Genotyping in replicates or controlling for Hardy-Weinberg equilibrium cannot identify misassignments caused by sample mix-ups. DNA-based sex identification (sex typing) is currently the best strategy to identify mix-ups. Here we review the available methods and present validated protocols for sex typing. The protocols are based on single-nucleotide differences between the human amelogenin genes, AMELX and AMELY, and are optimized for real-time PCR (TaqMan), primer-extension (SNaPshot) and PCR-RFLP genotyping platforms. In addition, we review the limitations of the sex-typing strategy, including a limited ability to identify single sample mix-ups, the dependence of the power of this approach on the sex distribution in the study population, and rare genetic conditions. Alternative strategies for mix-up identification and possible consequences of mix-up identification are also discussed.