Cementation of sandstones with quartz has a major influence on reservoir properties, but in many cases the sources for the quartz cement are not well constrained. Some of the suggested sources are internal, such as dissolution of feldspar or other silicate framework grain, or pressure solution of quartz; others are external sources supplied by large-scale flow systems or supply from the adjacent shales during compactional flow or diffusion. Many studies have suggested that the diagenesis of shales and interbedded sandstones is closely related and that shales act as a significant source for diagenetic quartz in the sandstones during diagenesis. Reported major silica sources in shale are; silica released from the conversion of a smectite to illite during burial diagenesis, pressure solution of detrital quartz and other silicates in deeply buried shales, dissolution of opaline marine skeletal grains (diatoms, radiolaria, sponge spicules), silica released from the hydration of volcanic glass, silica released detrital amorphous alumino-silicates that accumulated in muds, and decomposition of feldspars. Silica release in shale does not necessarily source sandstone diagenesis but may reprecipitate as opal, quartz or other mineral phases inside the shale itself. The deep marine sandstones in the Siri Canyon, Danish North Sea, have been reported to import significant amounts of dissolve silica from adjacent Paleocene shales during early diagenesis, and the authigenesis of silica developed several morphologies in the studied sandstones. We have studied the diagenesis of one of these shales, the Sele Formation shale, to document the diagenetic steps which release silica and to evaluate the possible capacity and timing of silica export from the shale into the interbedded sandstones. Initially, the shales were rich in smectite and had variable admixtures of silicious fossils (diatoms, radiolarian and sponge spicules) and volcanic ash. Depth dependant alteration of the various components in the shale, results in successive stages of silica-release. In shallow samples (<1700 m), the alteration of volcanic ash has already been completed. Released silica was partly consumed for the precipitation of smectite and zeolite. Opal-CT is not systematically related to volcanic ash, and some silica may have been mobilized and migrated into the interbeded sandstones. In addition, a major part of the biogenic silica has been transformed into opal-CT and partly to microcrystalline quartz. The microcrystalline quartz is an internal sink for dissolved silica, but the shale may also have been an active silica exporter during this transition. With deeper burial (2000-2900m), opal-CT is fully transformed to microcrystalline quartz. During this phase, silica has been partly mobile and depending on the rate of dissolution compared to the rate of precipitation, silica may have been lost to sandstone cementation. Zeolite is also dissolved and mobilized silica may also have activated the shale as silica exporter. At deep burial, iron-rich chlorite has replaced a minor part of smectite. The smectite to chlorite transformation released silica at the expense of iron. Therefore, a third phase of silica mobility was active. Microcrystalline quartz may have been an internal sink, but the shale would also be a potential silica supplier at this stage. At maximum burial depth (2900 m), the major part of the smectite is still not transformed, and the smectite to illite transformation has not yet been activated as a potential silica source. Early cementation by opal and microquartz depends on the supply of dissolved silica from a readily dissolvable source. The process starts with biogenic silica dissolution and supersaturation of pore fluids with respect to opal-CT and quartz. Relative rates of opal-A dissolution and opal-CT nucleation govern the extent to which silica activity is buffered and hence the possible silica flux into adjacent sandstones. With progressively less remaining opal-A, silica activity starts to decline and opal-CT nucleation becomes subordinate to growth. Precipitation of cryptocrystalline quartz began once the thermodynamic drive for the opal-CT became insignificant. Relative rates of opal-CT dissolution with respect to quartz nucleation and growth buffer the silica activity and the possible silica flux into adjacent sandstones. The early stages of the sandstone diagenesis are dominated by massive precipitation of opal cement and microquartz. This indicates a large flux of silica at the time of opal-A to opal-CT transformation and possibly at the time opal-CT to microquartz transformation. Therefore, the dissolution rate of precursor phases (biogenic silica, opal-A and/or opal-CT) was fast enough to create a sufficient silica gradient to exports silica from shale to sandstone.