Abstract: This study was aimed to estimate breed-specific genetic parameters of Large White pigs in Wen's Group and evaluate the accuracy of different methods of genomic selection of the trait of litter birth weight (LBW). Animal model and DMU software were used to estimate variances for the trait of LBW, including additive variance components and permanent environmental effect variance components, as well as heritabilities of the trait. The reproductive reference population for Large White pig genomic selection were built based on genotype-by-sequence (GBS) technology. The accuracy of genomic estimated breeding values (GEBV) were compared using different methods of BLUP, GBLUP and ssGBLUP in the validation group. Estimation of heritability was 0.08 for LBW, which was a trait with low heritability. The genetic correlation coefficients between LBW and some quantitative traits, such as total number born, number born alive, number born robust, number of weak, deformed, stillborn and mummified pigs, were 0.59 0.68, 0.88 and -0.17, respectively. According to the results of the genomic selection, the ssGBLUP had a higher accuracy of 0.38, increased 15.79% compared to the traditional BLUP method. The spearman rank correlation coefficient of GEBV between ssGBLUP and BLUP was 0.63. Selection of LBW has a positive effect on the improvement of litter size. The accuracy of genomic selection can be improved effectively combined with ssGBLUP method.