Flint is common in many successions, but since the nodules can be the result of a series of replacement and void filling processes, their origin and broader paleo-oceanographic significance can remain enigmatic. This study addresses the origin of flint, in the Campanian-Maastrichtian chalk of Denmark using the 450-m-long Stevns-1 core as a natural laboratory.Four types of siliceous nodules occur in the core: porcellanite, white flint, white flint with a core of dark flint, and dark flint. All nodules consist mainly of microquartz with subordinate lutecite, spherulitic chalcedony, megaquartz, and dolomite. The lutecite replaces macrofossil fragments, mainly bryozoans and shells, microquartz replaces coccoliths, and the chalcedony and megaquartz are pore-filling. The preservation of the external shape of bryozoans, shells, and coccoliths suggest that siliceous nodules were formed early by a one-by-one volume replacement of chalk. Values of d 18 O smow of microquartz and lutecite around 33% indicate temperatures of 15-17uC at the time of precipitation; in contrast, the d 18 O smow values of megaquartz indicate precipitation at temperatures up to 48uC. The occurrence of chalcedony together with more stable and ordered megaquartz in irregular voids indicates that silica gel and/or opal-CT transformed to the a-quartz phases, microquartz, lutecite, chalcedony, and megaquartz, through the dissolution-precipitation process (Ostwald ripening).These results motivate a conceptual model in which the replacement process was initiated by microbial decomposition of organic matter and took place during periods of low rate of, or even stopped, sedimentation which fixed the redox boundary and the microbial metabolic zones at a specific depth below the sea floor. Biogenic opal-A incorporated in the sediment column below the sea floor was dissolved and precipitated at the redox boundary as silica gel and/or opal-CT, possibly aided by sulfideoxidizing bacteria. During burial, the dissolution-precipitation process transformed the silica gel and opal-CT to the a-quartz phases.The results of this study indicate that precipitation of the flint precursor can be controlled by a combination of an abundant biosiliceous source and high microbial activity in the sediments immediately below the seafloor. Flint is therefore likely to reflect deposition in a narrow interval at relatively shallow depths and therefore might be useful in studies of shallow-water carbonates.