In this article, the authors aim to establish the suitable Newmark displacement models of the coseismic displacements for southwest China, where the 2008 Wenchuan earthquake, the 2013 Lushan earthquake, and the 2017 Jiuzhaigou earthquake occurred. Based on the 33 strong ground motion data recordings during the 2008 $Mw$ 7.9 Wenchuan earthquake, the Newmark displacement ($Dn$) value, the critical acceleration ($Ac$), and the Arias intensity ($IA$) value of the recordings are used for relationship analysis. The new parameters are then fitted to southwest China based on the functional form of Jibson’s models and their modified models. Results show that the models have higher correlation coefficients and smaller estimation errors. These are then tested by the strong ground motion data of the 2013 $Mw$ 6.6 Lushan earthquake, indicating that the models with new fitting parameters are suitable for the southwest area of China. According to comparison analysis, one of the models with newly fitted parameters has the smallest error and the maximum correlation coefficient: $logDn=0.465logIA+12.896AclogIA−22.201Ac+2.092±0.148(R2=0.92)$, being recommended for the southwest area of China. Subsequently, this model is used to evaluate the landslide hazards induced by the 2017 $Ms$ 7.0 Jiuzhaigou earthquake, the results of which are consistent with the survey results. Therefore, the recommended model is useful for the future earthquake hazard assessment in the southwest area of China. The results presented in this article also suggest that assessment models need to be refitted with new parameters for estimating the coseismic displacement in a specific region, which will get better estimated values.