The future wide-swath satellite altimeters, such as the upcoming Surface Water Ocean Topography (SWOT) mission, will provide instantaneous 2D measurements of sea level down to the spatial scale of O(10 km) for the first time. However, the validity of the geostrophic assumption for estimating surface currents from these instantaneous maps is not known a priori. In this study, we quantify the accuracy of geostrophy for the estimation of surface currents from a knowledge of instantaneous sea level using the hourly snapshots from a tide- and eddy-resolving global numerical simulation. Geostrophic balance is found to be the leading-order balance in frontal regions characterized by large kinetic energy, such as the western boundary currents and the Antarctic Circumpolar Current. Everywhere else, geostrophic approximation ceases to be a useful predictor of ocean velocity, which may result in significant high-frequency contamination of geostrophically computed velocities by fast variability (e.g., inertial and higher). As expected, the validity of geostrophy is shown to improve at low frequencies (typically$<$0.5 cpd). Global estimates of the horizontal momentum budget reveal that the tropical and mid-latitude regions where geostrophic balance fails are dominated by fast variability and turbulent stress divergence terms rather than higher-order geostrophic terms. These findings indicate that the estimation of velocity from geostrophy applied on SWOT instantaneous sea level maps may be challenging away from energetic areas.