Seismic observation has key advantages that, in the right geological conditions, make it extremely valuable for CO2 storage monitoring. Among surface techniques allowing the observation of the entire storage complex, it has by far the greatest vertical and lateral resolution and it can provide a full 3D view of the reservoir and overburden. However, in our experience, theoretical applicability of seismic acquisition for CO2 plume tracking has often been challenged by the geology: high rock stiffness, heterogeneity, large depths, or low porosity in a storage site are not favorable factors. Moreover, there is uncertainty on the minimum levels of detectability. CO2 might mix either homogeneously or create patches of variable saturation, and this can result in a large variability on the expected seismic signal, especially for saturations lower than 50%. We investigate the seismic signal expected for a location in the Gippsland Basin, Victoria, Australia. Using classical rock physics equations, CO2 detectability at the reservoir level will present some challenges, but the possibility for detection above the reservoir is quite favorable. An understanding of the relative heterogeneity of storage formations is critical to establishing uncertainty and detection limits of time-lapse seismic technology. (C) 2013 The Authors. Published by Elsevier Ltd.