Abstract High-resolution melting (HRM) shows great promise for high-throughput, rapid genotyping of individual polymorphic loci. We have developed HRM assays for genotyping single nucleotide polymorphisms (SNP) in several key genes that are involved in methyl metabolism and may directly or indirectly affect the methylation status of the DNA. The SNPs are in the 5,10-methylenetetrahydrofolate reductase (MTHFR; C677T and A1298C), methionine synthetase (MTR; 5-methyltetrahydrofolate-homocysteine methyltransferase; A2756G), and DNA methyltransferase 3b (DNMT3b; C46359T and C31721T) loci. The choice of short amplicons led to greater melting temperature (Tm) differences between the two homozygous genotypes, which allowed accurate genotyping without the use of probes or spiking with control DNA. In the case of MTHFR, there is a second rarer SNP (rs4846051) close to the A1298C SNP that may result in inaccurate genotyping. We masked this second SNP by placing the primer over it and choosing a base at the polymorphic position that was equally mismatched to both alleles. The HRM assays were done on HRM capable real-time PCR machines rather than stand-alone HRM machines. Monitoring the amplification allows ready identification of samples that may give rise to aberrant melting curves because of PCR abnormalities. We show that samples amplifying markedly late can give rise to shifted melting curves without alteration of shapes and potentially lead to misclassification of genotypes. In conclusion, rapid and high-throughput SNP analysis can be done with probe-free HRM if sufficient attention is paid to amplicon design and quality control to omit aberrantly amplifying samples. (Cancer Epidemiol Biomarkers Prev 2008;17(5):1240–7)