Investigated in this study is the response of a global ocean general circulation model to forcing provided by two wind products: operational analysis at 1000 mb from the National Center for Environmental Prediction (NCEP); observations made by the ERS-1 radar scatterometer. The sea level simulated by the model using the two wind fields is compared to the observations made by the TOPEX/POSEIDON radar altimeter for a period of two years. The focus of the analysis is placed on the large-scale ocean variabilities at mid and high latitudes. The sea level simulations resulting from the ERS-1 wind are found to be closer to the TOPEX/POSEIDON observations over most of the global oceans. However, after being scaled down in wind stress by 50%, the NCEP wind becomes consistent with the ERS-1 wind in driving the ocean model to simulate the seasonal and intraseasonal sea level variabilities. The sensitivity of the model to wind forcing at the investigated scales is particularly pronounced in the Southern Ocean, where the sea level variabilities are primarily caused by large-scale barotropic motions driven by wind. Other areas of appreciable sensitivity to wind forcing include the western and central North Pacific, the western subtropical South Pacific and the South Indian Ocean. The result of the study has demonstrated the sensitivity of a widely-used ocean model to the quality of wind forcing, as well as the synergistic use of two satellite sensors in the study of ocean dynamics.¿ 1997 American Geophysical Union