The salt fraction in permafrost soils/sediments of the McMurdo Dry Valleys (MDV) of Antarctica can be used as a proxy for cold desert geochemical processes and paleoclimate reconstruction. Previous analyses of the salt fraction in MDV permafrost soils have largely been conducted in coastal regions where permafrost soils are variably affected by aqueous processes and mixed inputs from marine and stratospheric sources. We expand upon this work by evaluating permafrost soil/sediments in University Valley, located in the ultraxerous zone where both liquid water transport and marine influences are minimal. We determined the abundances of Cl⁻, NO₃⁻, ClO₄⁻ and ClO₃⁻ in dry and ice-cemented soil/sediments, snow and glacier ice, and also characterized Cl⁻ and NO₃⁻isotopically. The data are not consistent with salt deposition in a sublimation till, nor with nuclear weapon testing fall-out, and instead point to a dominantly stratospheric source and to varying degrees of post depositional transformation depending on the substrate, from minimal alteration in bare soils to significant alteration (photodegradation and/or volatilization) in snow and glacier ice. Ionic abundances in the dry permafrost layer indicate limited vertical transport under the current climate conditions, likely due to percolation of snowmelt. Subtle changes in ClO₄⁻/NO₃⁻ ratios and NO₃⁻ isotopic composition with depth and location may reflect both transport related fractionation and depositional history. Low molar ratios of ClO₃⁻/ClO₄⁻ in surface soils compared to deposition and other arid systems suggest significant post depositional loss of ClO₃⁻, possibly due to reduction by iron minerals, which may have important implications for oxy-chlorine species on Mars. Salt accumulation varies with distance along the valley and apparent accumulation times based on multiple methods range from ∼10 to 30 kyr near the glacier to 70–200 kyr near the valley mouth. The relatively young age of the salts and relatively low and homogeneous anion concentrations in the ice-cemented sediments point to either a mechanism of recent salt removal, or to relatively modern permafrost soils (<1 million years). Together, our results show that near surface salts in University Valley serve as an end-member of stratospheric sources not subject to biological processes or extensive remobilization.