Flow-through experiments have been performed to study the thermodynamics of biogenic silica (opal) dissolution in deep-sea sediments. They were applied for the first time on sediment from the Southern Ocean and We have extended the use of these experiments to other deep-sea settings, thereby covering a wide range of in situ silicic acid asymptotic concentrations (Casympt; 200 to 900 μmol/L) and biogenic opal content (BSi; 0.5 to 80%). Performing these experiments under in situ bottom temperatures allows for the comparison between experimental apparent solubilities and Casympt concentrations. Low values of BSi apparent solubilities have been measured in the deepest sections of the multicores collected in the northeast Atlantic (229 μmol/L) and in the equatorial Pacific (505 μmol/L). They are only 10 to 20% higher than the in situ Casympt concentrations. This demonstrates a clear control of pore water silicic acid concentrations by the in situ apparent solubility of the BSi, i.e., the solubility of BSi within a complex sedimentary matrix that includes important quantities of silicate minerals.In regions where the percentage detrital/percentage biogenic ratio is low, the apparent solubility of the biogenic silica is close to that of in situ biogenic silica. In the opposite case, when the percentage detrital/percentage biogenic ratio is high, reprecipitation reactions induce strong interference on the dissolution properties of the opal, both in situ and in flow-through experiments. In such a sedimentary matrix, it is important to determine the appropriate opal solubility to be used in early diagenetic models, i.e., the solubility of the biogenic silica just before deposition on the seabed. This has been achieved by performing flow-through experiments on sediment trap material from the north Atlantic site. Comparison of apparent biogenic silica solubility measured by flow-through experiments and the silicic acid concentrations measured in the cups of the sediment traps suggested that the solubility of biogenic silica that reaches the sediment-water interface is not unique and varies spatially and temporally. In fact, it is the degree of coupling between surface waters and the sediment-water interface that will control the aging of biogenic silica in the water column and hence the dissolution properties of the biogenic silica deposited at the sediment-water interface. All these results call for a strong improvement of biogenic silica early diagenetic models that should include not only a reprecipitation term that takes into account interaction with silicate minerals but also the existence of several phases of biogenic silica and thus that should operate in a non-steady-state mode to account for seasonal variations in the quality of deposited biogenic silica.