The diurnal cycle in rainfall varies considerably from region to region in the tropics. Determining this variability is important both for comparing predictions of atmospheric models to real atmospheric behavior and for making sure that estimates of total rainfall from low-altitude satellites are not biased because of their infrequent observations of a given region of the earth. Although there are no data from the proposed Tropical Rainfall Measuring Mission (TRMM) satellite to work with yet, we can ask how well the diurnal cycle in rainfall will be detected when the satellite is eventually collecting data, given the satellite's proposed sampling characteristics. Data analyses for the diurnal cycle are discussed, taking into account the fact that the satellite visits will be irregularly spaced in time. The amplitudes of the first few harmonics will be determined by least-square fits to the satellite observations, and the tests needed to establish the statistical significance of the fitted amplitudes are discussed. The accuracy with which the first few harmonics of the diurnal cycle can be detected is estimated from several months of satellite data using rainfall statistics observed during the GARP (Global Atmospheric Research Program) Atlantic Tropical Experiment (GATE).