We present a new compilation of the structural and stratigraphic evolution of the Tharsis region of Mars that incorporates recent advances in understanding its stratigraphy, and we introduce a lithospheric deformation model that can account for the observations. The first period in the formation of Tharsis occurred in Late Noachian/Early Hesperian time with the deposition of volcanic plains materials throughout the surrounding highlands (e.g., Lunae Planum) and on the Tharsis rise (which includes the giant volcanoes and surrounding, elevated lava fields). Extensive radial normal faulting occurred on the rise, locally extending outward at Valles Marineris and Tempe Terra, and concentric wrinkle ridges formed along the edge of the rise. This regional deformation appears to have been modulated by a global compressional stress field due to rapid planetary cooling and contraction. The second period occurred during the Late Hesperian/Amazonian with Tharsis volcanism centered on the rise and radial extensional deformation that extended from the center of the rise for thousands of kilometers. We propose a model in which the lithosphere beneath Tharsis consists of a thin elastic crustal cap on the rise that is mechanically detached from the strong upper mantle by a volcanically thickened, hot, weak lower crust. These layers merge into a single cooler, strong lithospheric layer around the edges of the rise. This model is capable of generating large extensional hoop stresses throughout much of the western hemisphere, in agreement with observations. The tectonic interpretation of the stresses predicted by this model requires the reconciliation of extensional strain within narrow grabens and compressional strain within wrinkle ridges with (1) processes in the deeper lithosphere, (2) the sparsity of strike‐slip faults, and (3) other global or locally important stress fields. Stresses predicted by global models affect the entire thickness of the lithosphere, and they can be reconciled with narrow, closely spaced grabens that accommodate large amounts of extensional strain in the upper few kilometers of the lithosphere if the grabens are underlain and kinematically linked with dikes or other tension cracks, such as hydrofractures. Deeper levels of the lithosphere can accommodate this strain by elastic expansion if grabens are spaced far apart (many tens to hundreds of kilometers). Mechanical considerations suggest that deformation on faults beneath wrinkle ridges could extend through a significant thickness of the brittle crust. A number of factors, including stresses generated by the addition of overburden, intrusion of dikes, weakness of geologic materials under extension, and the laterally constrained nature of a single lithospheric plate, may have inhibited the formation of strike‐slip faults on Mars. Stresses generated from the removal of overburden could have augmented planetwide wrinkle ridge formation during the Late Noachian/Early Hesperian and in Kasei Valles and western Chryse Planitia during the Early Amazonian. The nonuniform distribution of tectonic features around Tharsis can be understood in terms of the concentration of regional stresses and strain near weaker volcanotectonic cent