Nonuniform rain rates within a field of view (FOV) and a nonlinear rain rate-microwave temperature (R-T) relation lead to a bias in the estimation of areal average rain rate from spaceborne microwave measurements. This bias is estimated from rain rate data collected during the Global Atmospheric Research Program Atlantic Tropical Experiment (GATE) using an R-T relation which is derived from the model results of Wilheit et al. (1977). The bias is about 25% (30%) for a footprint size of 8 km and increases to about 40% (45%) for a footprint size of 40 km for phase I (II) of GATE. In the large FOV limit of 280-km footprint size, the bias is 48% (50%). An experiment was performed in which the biases calculated from phase I were applied to phase II for different rain rate categoreis. The empirical correction works quite well. An approximate formula which takes account of the effect of spatial inhomogeneity within the FOV and a nonlinear R-T relation is derived. To first order, the two effects multiply to produce the bias. The bias formula is applied to rain field models. For a rain field model where the autocorrelation function is defined by an exponential, the dependence of the bias on the ratio of the FOV size and the e-folding scale is very similar to those calculated from the GATE data. For a Poisson process rain field model the bias formula shows an inverse dependence on the probability of rain. In this context, the lower percent bias for phase I of GATE can be understood in terms of the higher probability of rain during the same period. ¿ American Geophysical Union 1990