In recent years, TiO2, as a potential electrode material in Li and Na batteries, has been the subject of considerable experimental and computational research. However, the typical density functional theory (DFT) functionals used (e.g., the generalized gradient (GGA)) for such calculations are not without their shortcomings. To avoid these well-known issues, we report the first use of hybrid DFT calculations to calculate the Li and Na intercalation properties for anatase, rutile, and TiO2(B). The magnitude of GGA intercalation voltage underestimation is shown to vary depending on the polymorphs. We find that Li intercalation is most energetically preferred in anatase, while Na intercalation is most feasible for TiO2(B). Using the screened exchange hybrid functional, all intercalation processes are shown to be thermodynamically favorable, with the exception of Na in rutile. The electronic structures of these intercalated materials are also calculated, and significant improvements, in terms of band gap prediction and charge localization, are presented in comparison with GGA. We hope that our results will encourage more use of hybrid density functionals in the modeling of fundamental battery material properties.