Numerical simulation technology was used to study the filling stage of gas -water -assisted injection molding (GWAIM) process which is an innovative technology. The numerical simulation model of the filling process was established. The Cross-WLF viscosity model was used and the volume of fluid was adopted to track the free interfaces. The penetration process, temperature evolution, the effects of process parameters on the residual wall thickness of the GWAIM process and process optimization were systematically studied. The results showed that the residual wall thickness was uniform and the fluctuation range was small, and the temperature difference between the parts before gas injection and after water injection was large. The residual wall thickness tended to increase with the increase of gas injection delay time. The residual wall thickness of the product increased slightly and tended to be stable with the increase of water injection delay time. With the increase of gas injection speed and mold temperature, the residual wall thickness became smaller. With the increase of water injection velocity and melt injection temperature, the residual wall thickness of the parts decreased in a small range.Through orthogonal experimental range analysis, it is concluded that the gas injection delay time is the most important factor affecting the residual wall thickness of the parts, and the optimal combination of process parameters is gas injection delay time of 3 s, water injection delay time of 3 s, gas injection speed of 3 m/s, water injection speed of 2 m/s, mold temperature of 320 K and melt injection temperature of 503 K.