Abstract The stability of celestine (SrSO 4) in marine sediments has been investigated through the calculation of its saturation index at the in situ temperature and pressure, using the entire ODP/IODP porewater composition data base (14,416 samples recovered from sediments collected during 95 ODP and IODP Legs) that has been thoroughly corrected for missing data and inconsistencies. The porewater in situ pressure has been obtained from depth data, and the in situ sample temperature has been calculated from the bottom seawater temperature and from the measured thermal gradient. When the latter is unavailable, a default value of 35 °C/km or an assumed gradient similar to that of a site located close to the study area has been used. Molarities have been converted to molalities through the calculation of the porewater densities, which have represented by an empirical function of the total dissolved solids load up to 150 g/L. Ocean bottom waters are largely undersaturated with respect to celestine. Yet sediment porewater saturation is not uncommon: it is reached in 83 boreholes (i.e. about 10% of all boreholes) drilled during 22 ODP/IODP Legs. Celestine equilibrium can be reached through two different non-exclusive causes: a strontium increase in porewater linked to the dissolution of Sr-rich aragonite or biogenic calcite and precipitation of Sr-poor calcite (carbonate recrystallization), or an increase in the strontium and sulfate porewater concentrations linked to a salinity increase due to the presence of brines. Sediments at most of the sites exhibiting celestine equilibrium share common characteristics such as a carbonate-rich lithology (typically higher than 80 wt.%) and a low organic carbon content (generally below 1 wt.%). These results indicate that modification of porewaters during burial diagenesis can easily lead to celestine saturation, especially in carbonate-rich sediments. We then briefly discuss former interpretations of the presence of celestine in ancient and recent marine sediments, as well as the consequences of the incorporation of celestine formation in diagenetic models using the porewater Sr content and Sr isotopic composition.