Using hydrographic data collected by CTD during five cruises of the 1988 Black Sea Oceanographic Expedition, from 16 April to 29 July 1988, we describe the distribution of potential temperature (theta), salinity (S), and potential density (sigma(theta)) throughout the water column. The salinity and density increase rapidly with depth, while temperature decreases to a minimum at 50 m in the cold intermediate layer (CIL). All three variables increase slowly with depth in the deep water. The hydrographic properties of the upper 200 m varied little over the duration of the 1988 expedition. Significant differences are observed when the 1988 data are compared with the 1969 Atlantis II data set. All of the 1969 data are warmer at a given salinity than the 1988 data to a depth with a salinity of about 21.0 parts per thousand. Possible causes for these changes are increased heat loss to the atmosphere and decreased freshwater input. The most distinctive feature in the deep water is a homogeneous benthic bottom layer that extends from about 1700 m to the bottom. There is a single pronounced step in all hydrographic properties at the top of this layer. Vertical transport across the upper boundary may be controlled by double diffusion driven by geothermal heat flow. The predicted double-diffusive heat flux agrees with geothermal heat flow to within a factor of 5. A simple box model with surface, entrainment and deep-water reservoirs is used to model the entrainment process and the residence time of deep water in the Black Sea. The results suggest that the Bosporus inflow entrains water with properties of the CIL. The ratio of entrainment to Bosporus inflow is 3.3. Assuming a Bosporus inflow of 312 km3 y-1, the resulting residence time of the deep water is 387 years. A total CO2 balance is used to calculate the flux of carbon into the deep water and a C-14 balance is used to calculate the pre-nuclear value of DELTA-C-14 = -200 parts per thousand in the entrainment water. This highly depleted value would have resulted in an apparent age of 1400 years for the CIL and, probably, the surface water as well. If the carbon flux of biological origin was depleted to the same extent this may account for some of the differences in sedimentary chronology based on C-14 dates and varve counts.