At a reflective intensity modulated fiber optic microphone (RIM-FOM), the acoustic signal makes a membrane vibrate and modulate the reflected intensity. In the existing models of the RIM-FOM, the offset of all points of the membrane, due to the vibration, is assumed to be equal. However, this assumption does not represent the actual vibration of the membrane, which follows a continuous surface shape change caused by the acoustic signal. We establish a revised theoretical model in which the influence of the actual membrane surface shape change on the reflective intensity modulation is considered. Experiments show that there is a discrepancy between the experimental optimum operating distance and the analytical result from the existing model, while our new model gives a better agreement with the experimental results. In particular, our analysis shows that, in using the existing model, the other microphone performance characteristics are misestimated, while our revised model can provide a closer solution.