Based on ERA-Interim reanalysis during 1979–2014, the performance of 18 climate models from phase 6 of the Coupled Model Intercomparison Project (CMIP6), are assessed in terms of horizontal distribution of Pacific blocking frequency (PBF) throughout the wintertime and in the individual winter month, utilizing a 2D blocking index. It is found that all the selected models have a good simulation performance in the horizontal distribution of the PBF throughout the wintertime. Among the selected models, CanESM5 (Can), EC-Earth3 (EC), GFDL-CM4 (GFD-C), GFDL-EM4 (GFD-E), IPSL-CM6A-LR (IPS), IPSL-CM6A-LR-INCA (IPS-IN), MPI-ESM-1-2-HAM (MPI-HAM) and MPI-ESM-1-2-HR (MPI-HR) have better capabilities to simulate the entire wintertime PBF, and these eight models are chosen to be the key models. Further analyses show that the simulation performance of the PBF difference between January and December are better than that between January and February. From the assessment result of three objective measures (PCC, NRMSE, and TS) in January, in which month the Pacific blocking is the most active during wintertime, the multi-model-mean (MMM) of simulation performance in January is better than that in December and February. Among the key models, the MPI-HAM model ranks the top three of two measures (NRMSE and TS) in January. The GFD, IPS, and EC models rank the top three of one measure (TS for the GFD and EC, and NRMSE for the IPS) in January. But the EC model ranks the 13th in the other measure. Thus, MPI-HAM model is identified the best model on simulating the PBF during wintertime in 18 CMIP6 models, followed by the GFD and IPS models.