A simple model is presented for the calculation of relative signal-to-noise (S/N) ratios of coils of different sizes and configurations when applied to in vivo MRS. Axial symmetry is assumed, which enables rather simple expressions to be used for the calculation of coil loading by the tissue. The model is calibrated to experiments through measurement of the loaded and unloaded coil Q's. Applications of the model demonstrate that for small, superficial regions of interest (ROI), small surface coils can provide a S/N much improved over that of a larger coil. However, for very deep ROIs, larger coils or coils producing uniform B1 provide improved S/N.