In recent years, large numbers of cattle have been genotyped with SNP arrays from 3K to 800K. These platforms can be available to increase the efficiency and accuracy of breeding programs by implementing genomic selection. As for cattle, there are currently several imputation/phasing methods used in genomic selection, genome-wide association (GWA) studies, or genetic diversity analysis. Currently, many imputation and phasing methods are introduced to reduce the number of missing genotypes and to infer the haplotypes from these genotype data. Despite these efforts, imputation or phasing errors are still present. Next-generation sequencing (NGS) price has been consistently dropped, various population genomic theories and breeding program can be now applied to the sequencing data obtained from population of each breed of interest. For example, long-range haplotype sequencing technology can phase 99% of single-nucleotide variants (SNVs) in sequencing data without imputation process; current technologies typically phase ~95–97% in human. Therefore, we describe the phasing study using Hanwoo trio sample. First, we selected the representative trio sample from pedigree analysis in Hanwoo population. Genotyping was performed based on the Illumina 800K. Imputations for genotype data in this study were done using BEAGLE and FIMPUTE, genotype imputation tools that use a reference panel of haplotypes to estimate phase and impute missing genotypes in trio data. Second, We sequenced the trio data using Illumina Long-read haplotyping technology known as Moleculo. The short sequence reads producing from each molecule are assembled into synthetic long-reads. These fragments assign haplotype to homologous chromosomes in the phasing application. Finally, we compared accuracy of imputation/phasing based on the SNP array and sequencing data of an optimal reference panel of maternal/paternal haplotypes. These results help in improving selection and breeding value estimation and in avoiding imputation errors from SNP information.