The isotope composition of sulfate (SO42-) is an important geochemical tracer used to study elemental cycling globally. In polar desert environments such as Antarctica, sulfate in ice, snow, and surficial deposits has primarily been considered to derive from atmospheric deposition. However, recent studies suggest inputs from multiple sources may be important and possibly controlled by weathering, changing climate and seawater interactions in coastal areas. To better constrain the sources of SO42- in the McMurdo Dry Valleys (MDV), the largest of the ice-free regions in Antarctica, we investigated the delta S-34 and delta O-18 of SO42- in various depositional environments (lakes, ponds and valley bottoms) including archived Dry Valley Drilling Project sediment cores and modern surface materials. We found that, in general, the delta S-34 and delta(18)O of SO42- in the studied sediments, bedrock, and water samples exhibited wide ranges in values, consistent with multiple SO42- sources and post-depositional alterations in thaw zones of valley bottoms. For instance, the delta S-34 and delta O-18 of SO42- in Antarctic snow/ice and atmospheric aerosols generally had lower (+2 to +17 parts per thousand) and higher (-2 to +12 parts per thousand) values, respectively, compared to the higher delta S-34 (+3 to +71 parts per thousand) and lower delta O-18 (-25 to +1 parts per thousand) values in the MDV surficial sediments, bedrock, and lake deposits. This implies less SO42- contributions from atmospheric deposition that may also include local inputs from oxidation of H2S emitted from Antarctic lacustrine settings. Conversely, the distinctive delta S-34 ranging from +10 to +17 parts per thousand suggested varied inputs of SO42- from seawater (+21 parts per thousand) and weathering (oxidation) of bedrock- and lake-derived sulfide with low and high delta S-34 values (-2 to +3 parts per thousand and +5 to +14 parts per thousand, respectively). The highest SO42- concentrations (up to 3.5 wt% S) were measured in Lake Vanda sediments and were consistent with the presence of marine fossils and expected seawater-derived SO42- (median delta S-34 +22 parts per thousand) from past marine transgressions. Additionally, in many of the studied MDV lacustrine settings the delta S-34 and delta O-18 of SO42- showed significant increases (up to +71 and +7 parts per thousand, respectively) accompanied by distinctive high delta S-34 of sulfide/H2S (0 to +18 parts per thousand), suggesting microbial sulfate reduction in closed (ice-covered) lake systems. Furthermore, the negative delta O-18 values of SO42- (-20 to 0 parts per thousand) were considerably lower compared to seawater-SO42- (+9 parts per thousand) suggesting microbially- and/or abiotically driven oxygen isotope exchange between SO42-/SO32- and water, likely occurring on short and longer time scales. Overall, our results imply that without significant seawater inputs through marine transgressions to the MDV, the total SO42- deposition would be relatively small in lakes and valley bottoms due to low bedrock sulfide concentrations (<0.01 to 0.5 wt% S) and minor atmospheric deposition. Coupled isotope analyses and comparisons of the MDV materials used in this and previous studies enabled greater insight into assessment of variable atmospheric and geological sulfate sources and investigation of further isotope fingerprinting by microbial processes and marine inputs.