Abstract. It has been well established that during the late Quaternary, the Khvalynian transgression of the Caspian Sea occurred, when the sea level rose tens of meters above the present one. Here, we evaluate the physical feasibility of the hypothesis that the maximum phase of this extraordinary event (known as the “Early Khvalynian transgression”) could be initiated and maintained for several thousand years solely by hydroclimatic factors. The hypothesis is based on recent studies dating the highest sea level stage (well above +10 m a.s.l.) to the final period of deglaciation, 17–13 kyr BP, and studies estimating the contribution of the glacial waters in the sea level rise for this period as negligible. To evaluate the hypothesis put forward, we first applied the coupled ocean and sea-ice general circulation model driven by the climate model and estimated the equilibrium water inflow (irrespective of its origin) sufficient to maintain the sea level at the well-dated marks of the Early Khvalynian transgression as 400–470 km3/year. Secondly, we conducted an extensive 14C-dating of the large paleochannels (signs of high flow of atmospheric origin) located in the Volga basin and found that the period of their origin (17.5–14 ka BP) is almost identical to the recent dating of the main phase of the Early Khvalynian transgression. Water flow that could form these palaeochannels was earlier estimated for the ancient Volga River as 420 km3/year, i.e. close to the equilibrium runoff we determined. Thirdly, we applied a hydrological model forced by paleoclimate data to reveal physically consistent mechanisms of an extraordinarily high water inflow into the Caspian Sea in the absence of visible glacial meltwater effect. We found that the inflow could be caused by the spread of post-glacial permafrost in the Volga paleo-catchment. The numerical experiments demonstrated that the permafrost resulted in a sharp drop in infiltration into the frozen ground and reduced evaporation, which all together generated the Volga runoff during the Oldest Dryas, 17–14.8 kyr BP, up to 360 km3/year (i.e. the total inflow into the Caspian Sea could reach 450 km3/year). The closeness of the estimates of river inflow into the sea, obtained by three independent methods, in combination with the previously obtained results, gave us reason to conclude that the hypothesis put forward is physically consistent.