Understanding the end-Triassic mass extinction event (201.36 Ma) requires a clear insight into the stratigraphy of boundary sections, which allows for long-distance correlations and correct distinction of the sequence of events. However, even after the ratification of a Global Stratotype Section and Point, global correlations of TJB successions are hampered by the fact that many of the traditionally used fossil groups were severely affected by the crisis. Here, a new correlation of key TJB successions in Europe, U.S.A. and Peru, based on a combination of biotic (palynology and ammonites), geochemical (δ 13 C org ) and radiometric (U/Pb ages) constraints, is presented. This new correlation has an impact on the causality and temporal development during the end-Triassic event. It challenges the hitherto used standard correlation, which has formed the basis for a hypothesis that the extinction was caused by more or less instantaneous release of large quantities of light carbon (methane) to the atmosphere, with catastrophic global warming as a consequence. The new correlation instead advocates a more prolonged scenario with a series of feedback mechanisms, as it indicates that the bulk of the hitherto dated, high-titanium, quartz normalized volcanism of the Central Atlantic Magmatic Province (CAMP) preceded or was contemporaneous to the onset of the mass extinction. In addition, the maximum phase of the mass extinction, which affected both the terrestrial and marine ecosystems, was associated with a major regression and repeated, enhanced earthquake activity in Europe. A subsequent transgression resulted in the formation of hiati or condensed successions in many areas in Europe. Later phases of volcanic activity of the CAMP, producing low titanium, quartz normalized and high-iron, quartz normalized basaltic rocks, continued close to the first occurrence of Jurassic ammonites and the defined TJB. During this time the terrestrial ecosystem had begun to recover, but the marine ecosystem remained disturbed.

Sedimentology and sampling: Sedimentological logs were measured in the N Albert quarry in southern Sweden and in cored sections from deep wells in the Stenlille area in eastern Denmark, as well as the Rødby-1 core in southern Denmark (Figs. 2, DR1-DR2). Similarly, the Schandelah and Mariental cores from Germany, the Grouft core and a temporary road construction outcrop Junglinster Heedhaff in Luxembourg, were measured and logged (Figs. 2, DR3). Drillcore sampling of the Luxembourg material was carried out at the Service Géologique Luxembourg and laboratory analyses at the Steinmann-Institute (University of Bonn). The interpretation of the depositional environments is supported by detailed studies of the palynology, coal petrology, stable isotope geochemistry and wire-line log motifs (