Spring deposits reveal the timing and environment of past groundwater discharge. Herein, however, the potential for fossil spring deposits to infer water sources and palaeoflowpaths through trace elements and stable and radiogenic isotopes is examined.Past discharge (70 to 285 ka) in the Tecopa Basin in the Death Valley region of southeastern California is represented by tufa deposits, including mounds, pools, cemented ledges and rare calcite feeder veins. d 18 O values indicate that spring discharge was a mixture of far-travelled (regional) water with a significant, and perhaps dominant contribution of local recharge on a nearby range front and alluvial pediment, rather than simply representing an elevated regional water table. d 13 C values indicate regional water had a high TDS, whereas solute data imply low overall solute contents, consistent with dilution by a large component of local recharge.Radiogenic isotope data (U-series, 87 Sr/ 86 Sr) for tufa indicate that siliciclastic rocks (a regional aquitard) interacted with discharging water. To access this aquitard, regional flow was probably partitioned into a permeable north-south damage zone of a north-south range-bounding fault along the foot of the Resting Spring Range, which ultimately controlled the location of groundwater discharge. Existing models for modern discharge in the Tecopa Basin, by contrast, call upon westward interbasin flow in carbonate rocks from the Spring Mountains through the intervening (and nearly perpendicular) Nopah and Resting Spring Ranges. Understanding the controls on regional groundwater flow is critical in this and other arid regions where water is, by definition, a scarce resource. Thus, although it is a case study, this report highlights a fruitful approach to palaeohydrology that can be widely applied.