Capturing carbon dioxide (CO2) from power plants and other sources of its emissions and injecting it into deep geologic formations for permanent storage purposes has become an important area of research recently. Some of the good formation candidates for CO2 sequestration are deep saline aquifers because they have a high storage capacity and because they are available in almost all sedimentary basins of the world. The effectiveness of the sequestration process in saline aquifers depends highly on the two-phase CO2-brine relative permeability behavior and the latter can only be measured in the laboratory. Although relative permeability data for hydrocarbon bearing formations are generally available, very little is known about the relative permeability of CO2-brine two phase systems, especially for the cases involving actual reservoir rocks. In this paper, we present relative permeability data generated on three saline formations that are being explored as potential candidates for CO2 capture and sequestration. These are - the Knox sandstone, the St. Peter sandstone, and the Sylvania formation. The latter is divided into an upper sandstone formation and a lower dolomite formation, and both are investigated separately. Three 1.5 inch diameter by 2 inch long core plugs from each formation, drilled in the direction of flow, were used to form composite cores in order to conduct drainage CO2-brine relative permeability measurements using the unsteady state method at simulated reservoir conditions. In order to understand the various phenomena taking place during multiphase displacement processes, all of these experiments were conducted inside an X-ray CT (computerized tomography) scanner. Pixel-by-pixel image subtraction was used to calculate individual saturations of brine and CO2 at various stages of displacement. CT-based heterogeneity measurements were used to compare the heterogeneity of different plugs and to assess the effect of heterogeneity on displacement. Additionally, the CT-based image data were used to see if capillary end effects, gravity segregation and bypassing occurred during each test. The results indicate low CO2 end point relative permeability (k rg) in the range of 0.036 to 0.34 for the three formations. These low values can be attributed to rock heterogeneity as indicated by the saturation distribution determined from the scanner images. Such low relative permeability values would tend to decrease injectivity and hence increase the displacement efficiency. This was reflected on the residual brine saturations which were in the range of 0.25 to 0.45 indicating good storage potential for the investigated formations. This paper also includes some lab-scale numerical simulation runs which were conducted for history matching of the laboratory data and to have better confidence on the relative permeability data. The three-dimensional grid (50×50×120) was developed using the actual porosity data generated from the CT-scanning of the core plugs at 1.25 mm axial resolution. The porosity data were converted to permeability values using the Kozeny-Carman equation and the average permeability derived this way for each core plug was compared against the value measured individually using the brine permeability based data. For numerical simulation the laboratory conditions used for the relative permeability tests were used and ECLIPSE simulator with laboratory units was used to describe the cores in three dimensions. The results showed a good agreement between the lab-derived and simulator generated brine recovery curves and pressure drop between the inlet and outlet ends giving confidence in the process.