Marine-evaporitic brines frequently display Na, Cl and Br concentrations that significantly deviate from seawater evaporation paths, yielding markedly conflicting degrees of evaporation calculated for a specific brine. Here we present 493 new and 33 previously reported analyses of Ca-chloridic waters of Neogene age from the Dead Sea Rift (DSR) valley to explain such offsets. The DSR brines plot along an almost perfect mixing line (R 2 = 0.990) on a Br/Cl-Na/Cl diagram, extending between two end members A and B. Points A and B are located at Na/Cl = 0.804 and Br/Cl = 0.00193, and at Na/Cl = 0.00773 and Br/Cl = 0.0155, respectively, within the halite and bischofite stability fields. Brines A and B originated in a dual-mode evaporation basin. Brine A formed under the classic lagoon scenario (mode A), with seawater inflow and brine outflow at steady state. Occasional drops in water level, imposed by climatic or tectonic causes, resulted in outflow cutoff and in rapid concentration buildup. The second mode (B) initiated upon equilibration of the activity of water in the brine with the overlying relative humidity, resulting in composition and salinity approaching that of brine B, sustaining it until the next reversal to mode A. Thick evaporite deposits inhibited infiltration of brines A and B into the subsurface terrain, a process that was enabled only when the brine reached the permeable carbonate rock rim and border faults of the basin. Hence, brines that formed during the relatively short shifts from mode A to mode B could not penetrate into the deep subsurface, and bittern minerals that were formed during the frequent mode shifts were dissolved and flushed out into the sea upon the next resumption of outflow. The proposed model accounts for the deviations of brines from the marine evaporitic evolution curve by brine mixing, rather than due to a change in ocean chemistry. It also explains the absence of bittern minerals in the thick halite and gypsum/anhydrite succession, and the compositional gap between the widely different end member hypersaline fluids. This model applies directly to the studied DSR brines and evaporites, but it may be relevant to other inland evaporitic basins.