Recent advances in genotyping technologies enable us to rapidly generate genome-wide genotyping data. For example, a high-density SNP array can produce hundreds of thousands of genotype calls simultaneously. This technology is useful for population genetics studies such as diversity analyses and association mapping. However, when an array is applied to a biparental population, the cost-effectiveness is quite low because it inevitably contains many probes without polymorphisms between parents. With the advent of next-generation sequencing technology, a genotyping-by-sequencing approach becomes one of the options for genotyping of multiple samples. However, due to a biased distribution of polymorphisms and a large volume of missing data, it is not always a suitable approach for constructing genetic maps of wheat. To overcome these problems, we proposed a target amplicon sequencing procedure using three bi-parental doubled haploid populations. The populations consist of 188 lines each and share the paternal variety 'Kitahonami' which is a leading Japanese variety with high yield and superior quality. We previously surveyed DNA polymorphisms among Japanese varieties including the four parents by gene-targeting sequence capture and developed more than 3,000 markers for amplicon sequencing. Taking advantage of the Chinese Spring reference sequences (IWGSC RefSeqV1), we selected around 500 markers for each population that showed polymorphisms between the parents and were evenly distributed across the genomes. Primers of selected markers were pooled into a single tube, and multiplex PCR was performed using the primer mix. By indexing samples, genotypes of 188 doubled haploid lines were determined by a single run of the next-generation sequencer. Along with genotypes of SSR and functional markers, genetic maps consisted of 555, 685 and 683 markers with the total lengths of 3920.5, 3493.2 and 4718.0 cM, respectively. Even though the number of mapped markers was relatively small, physical positions of the markers revealed that cumulative sizes of these maps reached to 13.27, 13.51 and 13.24 Gb. They correspond to 94.4, 96.1 and 94.1 % of the reference genome size (14.07 Gb), respectively. The D genome maps, which tend to be low coverage due to their low polymorphic nature, covered 3.75 (94.8 %), 3.79 (95.9 %) and 3.74 Gb (94.6 %) of genomic region. Therefore, the procedure provided us with saturated genetic maps with a short turn-around time and low cost. Using these maps, we showed the usefulness of meta-QTL analysis for detecting genetic factors that are valuable for breeding.