Accurate modeling of the spatio-temporal cross-correlation between the subchannels of a multiple-input multiple-output (MIMO) channel is an important prerequisite of multi-element antenna system design. In this thesis, a new model for indoor MIMO channels is proposed, and a closed form expression for the spatio-temporal cross-correlation function is derived. This new analytical correlation expression includes many physical parameters of interest such as the angle-of-arrivals at the base station and the user, the associated angle spreads, and other parameters, in a compact form. Comparison of this model with narrowband indoor MIMO data collected at Brigham Young University exhibits the utility of the model. Specifically, capacity calculations and the application of the model to maximum likelihood detection in correlated narrowband MIMO channels demonstrates close match to empirical data. As a different approach to indoor correlation modeling, the commonly used Kronecker product model is considered, which shows large deviation from the measured data in terms of correlation, capacity, and bit error rate.