Permafrost-related processes are key ecosystem drivers in the Arctic system and often an indicator for long-term environmental change. Understanding past periods of permafrost degradation and aggradation is crucial to estimate future response of the terrestrial Arctic to climate change. We collected sediment cores from one thermokarst lake and two drained thermokarst lake basins from the northern Seward Peninsula and the Teshekpuk Lake region (Alaska) to gain insights into past landscape dynamics since the late Pleistocene in these continuous permafrost regions. We applied a multi-proxy approach on sediment cores using methods of micropaleontology (ostracods, rhizopods), sedimentology (grain size analyzes, magnetic susceptibility), biogeochemistry (TN, TC, TOC, δ13C), as well as geochronology (AMS radiocarbon dating, tephrochronology). The environmental evolution of the three study sites covers different temporal scales. Preliminary radiocarbon dating results of three short cores (core ID P_1, P_2 and P_3) from a thermokarst lake North of Teshekpuk Lake revealed mid- to late Holocene ages with numerous age reversals in spite of well layered lake sediments. A core-based study (core ID Kit-43) of thermokarst lake sediments on northern Seward Peninsula yielded a radiocarbon chronology in agreement with stratigraphic sequences; with its base dated to the Last Glacial Maximum at 22,800 cal a BP. The third core (core ID Kit-64) from a drained lake basin in the same study region was dated to the Early to Mid-Wisconsin and will be discussed in detail. As observed in aerial and satellite images, GG basin on the northern Seward Peninsula drained in Spring 2005 (Figure 1, Lenz et al., accepted). The core recovered in spring 2009 has preserved the permafrost re-aggradation after lake drainage with frozen sediment from the lake sediment surface down to a depth of 266 cm (Figure 2). Below 266 cm the unfrozen talik was still existent. The sedimentary record yielded prevailing terrestrial conditions prior to 45,000 a BP. Fine grained, cryoturbated, organic-rich Yedoma deposits indicated harsh and cold climate conditions in central Beringia during the Early to Mid-Wisconsin. An intermediate peaty layer with hydrophilic rhizopods signaled wet conditions by about 44,500 to 41,500 a BP. This potential initial thermokarst development was interrupted by deposition of a 1-m-thick layer of tephra which could be associated with the South Killeak Maar eruption by about 42,000 a BP. The wetland did not re-initiate after the tephra fall-out but terrestrial Yedoma deposition dominated the late Mid-Wisconsin. A sedimentary hiatus was dated to 23,000 to 300 a BP. Here, either a lack of deposition occurred between the Late Wisconsin until the late Holocene or thermokarst-related erosion may have dominated at the study site. In the latter case, the Yedoma upland of the modern GG basin has served as sediment source and Late Wisconsin deposits were eroded. The lake forming GG basin evidently initiated about 300 a BP as indicated by distinct lamination, freshwater ostracods (e.g. Fabaeformiscandina protzi) and hydrophilic rhizopods (e.g. Cyclopyxis kahli). Our investigation demonstrates that lake development in the permafrost-affected terrestrial Arctic can be triggered but also interrupted by global climate change (e.g. syngenetic permafrost formation during the Early to Late Wisconsin, rapid warming and wetting in the early Holocene), regional environmental dynamics (e.g. nearby volcanic eruptions and tephra deposition) or local disturbance processes (e.g. lake drainage). The present study emphasizes not only that permafrost formation as well as degradation in central Beringia was influenced by processes of local to global scale but also highlights that Arctic lake systems and periglacial landscapes are highly dynamic.