Great advances have been made recently in searching for individual identification single-nucleotide polymorphisms (IISNPs or IDSNPs). Such SNPs as suggested by SNPforID scientists and by Pakstis et al., are promising, although they were selected from older or smaller databases rather than the most recent database. Here, we describe a new computational strategy for developing IDSNPs based on HapMap. We searched through HapMap r27 for SNPs having minor allele frequencies ≥0.30 in all its 11 populations and found more than 1881 qualified SNPs. We examined 96 of them with 183 DNA samples from three Chinese populations using Illumina arrays. The average allele frequency for these 96 SNPs among the three populations was 0.495/0.505, the average number of identical SNP genotypes shared by two individuals among the 14 populations (three Chinese and 11 HapMap) was 37.9, and the random matching probability for two unrelated Hans to match in all 96 genotypes was 9.793 × 10(-39). Thus, most of these 96 SNPs are universally applicable.