The performance of GAMIL 1.0, a grid atmospheric model developed by the State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics, in simulating the western North Pacific summer monsoon is investigated. Influence on the simulation exerted by the inclusion of wind gustiness parameterization is discussed. Results show that major features of both the mean state and interannual variability of WNPSM are reasonably reproduced in the original version, but the strength is underestimated. After the inclusion of wind gustiness parameterization, improvements are obvious and mainly displayed as follows: the regional mean precipitation increases from 5.71 mm/d to 8.35 mm/d, which is more close to the observation; the correlation coefficient between the simulated and observed monsoon indices increases from 0.66 to 0.82; the amplitude of interannual variability becomes larger; more important, the simulated “ENSO－monsoon” relationship is almost the same as the observation. Analyses on the latent heat flux, heating rate, and divergent circulation indicate that the improvement in the mean state should be attributed to the enhancement of surface latent heating over the western North Pacific region. The improvement of interannual variability is related to the better mean state. The inclusion of wind gustiness parameterization leads to stronger surface wind speed, which enhances the surface latent heating. Then the mean precipitation is increased, indicating that the climate mean precipitation is improved. Following the improvement of climatological mean precipitation, the model's response to El Niño-type SST forcing is significantly improved, which is evident for both the heating field and corresponding upper-tropospheric convergence center in the western North Pacific.