Sampling error has long been an important issue in satellite-based rainfall measuring mission planning such as for TRMM and GPM (Global Precipitation Measurement), and in evaluating global maps of space-time average rain rate from multi-satellite rainfall data. Sampling errors due to sparse space-time rainfall observations from five currently operated low-earth-orbiting satellites (Aqua, DMSP-F13, F14, F15, TRMM (Tropical Rainfall Measuring Mission)) were evaluated using 3-year radar-AMeDAS composite data around Japan. We simulate realistic observational patterns for the satellite-borne microwave radiometers. Root-mean-square sampling error is expressed as a fraction of true average rain rate. We formulate sampling error as a function of space scale ranging from 0.1¿ to 5.0¿, time scale from 1 to 30 (day), and true average rain rate from 0.001 to 15 (mm/h) for five different combinations of these five satellites.