The Tecopa basin in eastern California was a terminal basin that episodically held lakes during most of the Quaternary until the basin and its modern stream, the Amargosa River, became tributary to Death Valley. Although long studied for its sedimentology, diagenesis, and paleomagnetism, the basin’s lacustrine and paleoclimate history has not been well understood, and conflicting interpretations exist concerning the relations of Tecopa basin to the Amargosa River and to pluvial Lake Manly in Death Valley. Previous studies also did not recognize basinwide tectonic effects on lake-level history. In this study, we focused on: (1) establishing a chronology of shoreline deposits, as the primary indicator of lake-level history, utilizing well-known ash beds and new uranium-series and luminescence dating; (2) using ostracodes as indicators of water chemistry and water source(s); and (3) correlating lake transgressions to well-preserved fluvial-deltaic sequences. During the early Pleistocene, the Tecopa basin hosted small shallow lakes primarily fed by low-alkalinity water sourced mainly from runoff and (or) a groundwater source chemically unlike the modern springs. The first lake that filled the basin occurred just prior and up to the eruption of the 765 ka Bishop ash during marine isotope stage (MIS) 19; this lake heralded the arrival of the Amargosa River, delivering high-alkalinity water. Two subsequent lake cycles, coeval with MIS 16 (leading up to eruption of 631 ka Lava Creek B ash) and MIS 14 and (or) MIS 12, are marked by prominent accumulations of nearshore and beach deposits. The timing of the youngest of these three lakes, the High lake, is constrained by a uranium-series age of ca. 580 ± 120 ka on tufa-cemented beach gravel and by estimates from sedimentation rates. Highstand deposits of the Lava Creek and High lakes at the north end of the basin are stratigraphically tied to distinct sequences of fluvial-deltaic deposits fed by alkaline waters of the Amargosa River. The High lake reached the highest level achieved in the Tecopa basin, and it may have briefly discharged southward but did not significantly erode its threshold. The High lake was followed by a long hiatus of as much as 300 k.y., during which there is evidence for alluvial, eolian, and groundwater-discharge deposition, but no lakes. We attribute this hiatus, as have others, to blockage of the Amargosa River by an alluvial fan upstream near Eagle Mountain. A final lake, the Terminal lake, formed when the river once again flowed south into Tecopa basin, but it was likely short-lived due to rapid incision of the former threshold south of Tecopa. Deposits of the Terminal lake are inset below, and are locally unconformable on, deposits of the High lake and the nonlacustrine deposits of the hiatus. The Terminal lake reached its highstand at ca. 185 ± 21 ka, as dated by infrared-stimulated luminescence on feldspar in beach sand, a time coincident with perennial lake mud and alkaline-tolerant ostracodes in the Badwater core of Lake Manly during MIS 6. A period of stillstand occurred as the Terminal lake drained when the incising river encountered resistant Stirling Quartzite near the head of present-day Amargosa Canyon. Our studies significantly revise the lacustrine and drainage history of the Tecopa basin, show that the MIS 6 highstand was not the largest lake in the basin as previously published (with implications for potential nuclear waste storage at Yucca Mountain, Nevada), and provide evidence from shoreline elevations for ∼20 m of tectonic uplift in the northern part of the basin across an ENE-trending monoclinal flexure.