The whiteleg shrimp, Litopenaeus vannamei, is an introduced species in China; the genetic relationship and genetic background among introduced populations in China are not clear. To analyze the effects of these unknown parent groups (UPG) in the basic population of L. vannamei, three populations with different growth rates and survival rates were collected as founder populations and were established as the base population by the diallel-cross design. Four models, namely, the Best Linear Unbiased Prediction based on Pedigree (pBLUP), the pBLUP with genetic groups (pBLUP-GG), the Single Step Genomic BLUP (ssGBLUP), and the ssGBLUP with Metafounders (ssGBLUP-MF), were used to estimate variance components for the body weight of the base population. The effects of UPG on genetic parameters and breeding value across the four models were compared and analyzed by cross validation. The results showed that the highest additive genetic variance (22.58 ± 4.39) and heritability (0.91 ± 0.10) were obtained using the pBLUP model. The pBLUP-GG, ssGBLUP, and ssGBLUP-MF models decreased additive genetic variance by 16.25%–61.20% and heritability by 15.38%–46.15% compared with that of the pBLUP model. The 5-fold cross validation for genotyped and ungenotyped individuals indicated that the accuracy of the pBLUP model was the lowest (0.64–0.68) of all estimates, and the accuracies of the other three models increased by 7.25%–10.53% compared with that of the pBLUP model. The prediction bias of the other three models decreased by 2.83%–7.56% compared with that of the pBLUP model (1.06–1.19). In conclusion, the models considering UPG effects could avoid the overestimation of additive genetic variance components for body weight in the basic population of L. vannamei.