Precipitation is the single most important determinant of the fluxes and states of the land surface hydrological system and the most important atmospheric input to hydrological models. Satellite-based precipitation estimates, such as those anticipated from the Global Precipitation Measurement (GPM) satellites, hold great promise for application in hydrologic simulation and prediction, especially in parts of the world where surface observation networks are sparse. However, the usefulness of these precipitation products for hydrological applications will depend on their error characteristics. Of particular interest in satellite-derived precipitation estimates is the sampling error, that is, the error in accumulated precipitation due to periodic sampling of the precipitation rate. To assess the effect of this error on simulated hydrological fluxes and states, synthetic error fields were imposed on an observation-based 1/2¿ latitude/longitude gridded precipitation data set. In turn, the generated precipitation fields were used as input to a macroscale hydrology model (MHM). Our results show that (1) streamflow errors were large for small drainage areas but decreased rapidly for drainage areas larger than about 50,000 km2. Much of the streamflow error is associated with fast (near-surface) runoff response. (2) Streamflow estimates were biased upward due to sampling errors, with the bias increasing with sampling interval and with drainage area. Evapotranspiration was biased downward in a compensating amount. (3) Spatial correlation of precipitation errors reduced the rate at which errors decreased with drainage area for all variables investigated, but the differences between the correlated and uncorrelated error cases were smaller for streamflow and evapotranspiration than for precipitation.