High-resolution melting (HRM) has been considered as a fast and simple single-nucleotide polymorphism (SNP) scanning and genotyping method for identifying differences in the shapes of melting curves between different genotypes. Fifty-six SNPs were developed in the Pacific oyster, Crassostrea gigas by mining expressed sequence tags database, using the HRM method. The frequency of the SNPs was estimated at 1 per 113 bp of contig sequences. Analysis of segregation in a full-sib family showed that 28 SNPs were polymorphic, with 15 in accordance to expected Mendelian ratios. Linkage grouping of the 28 markers resulted in six linkage groups. The combined power of exclusion of the 42 SNPs, which conformed to Hardy–Weinberg equilibrium, was greater than 99.98%, while the average polymorphic information content was 0.2223. The simulation results showed that the success rate of parentage analysis could be 97% with the 42 SNPs. These SNPs will be useful for pedigree analysis, association studies, and marker-associated selection of this species.