An attempt has been made to identify the processes that give rise to a number of depth trends in the magnetization of a 3.1 km vertical section of Icelandic-type oceanic crust and to assess the possibility that similar processes act, and depth trends occur, in typical oceanic crust. The depth trends in the Icelandic section consist of a general increase in saturation and induced magnetization to 2 km crustal depth, below which flow magnetization decreases while dike magnetization remains constant, and of large changes in flow magnetization that occur on a scale of a few hundred metres below 3 km crustal depth.Increase in saturation and induced magnetization with depth in the upper 2 km is thought to be the result of two processes: a decrease in low-temperature oxidation from the original lava surface to 700–800 m crustal depth, thence an increase in hydrothermal alteration with depth. This interpretation is based on oxide petrography and Curie temperatures, which show a weakly defined minimum in the 700–800 m interval, then an increase to ubiquitous "magnetite" values at just below 2 km crustal depth. Although the relationship between magnetic properties and oxide alteration is reasonably well known for the low-temperature oxidation process from laboratory studies and ophiolite and typical ocean-crust analogs, the change in magnetic properties during hydrothermal alteration is not generally known, nor are ophiolite or typical ocean-crust analogs presently available.Decrease in flow saturation and induced magnetization below 2 km is likely to be the result of alteration of magnetite (sensu lato) to nonmagnetic phases, either on a fine scale to hematite (s.l.) between 2 km and 3 km, or by leaching of iron, leaving anatase pseudomorphs after magnetite (s.l.) below 3 km. The relatively low porosity of the dikes is likely to have protected dike magnetite below 2 km from such oxidation and leaching processes.The study confirms that secondary magnetite in several forms is an important magnetic constituent of the flows in the lower part of the section, particularly where decomposition of primary magnetite is widespread. Secondary magnetite occurs as vermiform or bladelike masses, as rims associated with former silicates, or as fresh continuous magnetite occurring either as subhedral grains or as "reconstructed" primary grains in which relics of sphene-replaced ilmenite lamellae grids are seen.In conclusion, the possibility that the near-surface magnetization of typical ocean crust is commonly the minimum value for a layer extending downwards to the onset of an epidote-bearing facies deserves serious consideration, as does the possibility that strong, stable magnetization of secondary origin occurs in flows where dike density becomes significant.