Sequestration/enhanced oil recovery (EOR) petroleum reservoirs have relatively thin injection intervals with multiple fluid components (oil, hydrocarbon gas, brine, and carbon dioxide, or [Formula: see text]), whereas brine formations usually have much thicker injection intervals and only two components (brine and [Formula: see text]). Coal formations undergoing methane extraction tend to be thin [Formula: see text] but shallow compared to either EOR or brine formations. Injecting [Formula: see text] into an oil reservoir decreases the bulk density in the reservoir. The spatial pattern of the change in the vertical component of gravity [Formula: see text] is correlated directly with the net change in reservoir density. Furthermore, time-lapse changes in the borehole [Formula: see text] clearly identify the vertical section of the reservoir where fluid saturations are changing. The [Formula: see text]-brine front, on the order of [Formula: see text] within a [Formula: see text]-thick brine formation at [Formula: see text] depth with 30% [Formula: see text] and 70% brine saturations, respectively, produced a [Formula: see text] surface gravity anomaly. Such an anomaly would be detectable in the field. The amount of [Formula: see text] in a coal-bed methane scenario did not produce a large enough surface gravity response; however, we would expect that for an industrial-size injection, the surface gravity response would be measurable. Gravity inversions in all three scenarios illustrate that the general position of density changes caused by [Formula: see text] can be recovered but not the absolute value of the change. Analysis of the spatial resolution and detectability limits shows that gravity measurements could, under certain circumstances, be used as a lower-cost alternative to seismic measurements.