We explore the rainfall effect on the capacity and achievable rates of millimeter-wave (mmW) multiple-input multiple-output (MIMO) systems. We first develop a tractable channel model for point-to-point millimeter-wave (mmW) MIMO systems to characterize the rainfall effect. The model development is based on the known stochastic properties of signal propagation in a random medium. Under this model, the channel capacity of the mmW MIMO system at various rain rates is evaluated. We show that rain does not always have a negative impact on the system performance, provided that accurate instantaneous channel state information (CSI) is available at both the transmitter and receiver. We also show that by exploiting statistic channel information at the transmitter, the statistic water-filling (SWF) strategy allows the mmW MIMO system to approach near-optimal performance.