The measured magnetic properties of submarine igneous rocks, comprising data from approximately 300 specimens, are summarized. Basaltic rocks dominate the collection numerically, and are distinguished by their high Q (ratio of remanent to induced magnetic intensities). Limited numbers of altered samples indicate that spilitization, chloritization, and serpentinization can drastically reduce the intensity of magnetization. The available thermomagnetic data suggest that low Curie points may be typical of quenched basalts. The limited range of submarine igneous rock types examined, and the strong bias towards quenched samples necessitates a supplement to this summary in the form of a discussion of studies of magnetic properties from selected igneous rocks outcropping above sea level. In these studies, serpentinization of ultrabasic rocks has been observed in one case to increase the intensity of magnetization; chloritization and spilitization are confirmed as being magnetically destructive; maghaemitization may have destructive effects; titanomagnetite oxidation variation dominates in magnetic change of basaltic lavas (and some corresponding chemical changes are likely to occur); basaltic intrusives have a much more limited titanomagnetic oxidation range than is generally observed in lavas; and spontaneous demagnetization with time probably exists, at least in basalts. New data are presented. These include the magnetic properties of harzburgites dredged from the Macquarie Ridge, and eight pillow basalts from the South Pacific and Scotia Sea. The former suggest that harzburgite is capable of creating strong magnetic anomalies. Samples for the latter study were sufficiently large for study of the variation of magnetic and petrological properties with depth beneath the cooling surface. Systematic texturual changes from glassy exterior, through a variolitic zone to aphanitic interior characterize the silicates in most samples. Chloritization is present in some aphanitic parts. Serpentinization is present in some aphanitic zones and also next to joints. The opaque minerals were studied in detail in one pillow. The titanomagnetites are all fine and of low oxidation state. Very fine sulphides are common. The intensity of magnetization and suceptibility variation are closely related to the changes in titanomagnetite grain size. Although optically undetectable in the titanomagnetites, a zone of slightly higher oxidation is inferred to exist towards the centre of the pillow by the presence of higher Curie points and magnetic stability, and lower sulphide content. New data are also presented from traverses of Icelandic lavas and dykes, and from spilites of St Thomas, Virgin Islands. It is concluded that the submarine basalt magnetic properties which have so far been determined are largely a function of quenching, in contrast with the data from lavas outcropping above sea level which have generally experienced longer cooling periods, and which therefore include a greater range of titano-magnetite grain size and oxidation states. The quenching process can apparently proceed faster than the oxidizing process in basalts. Magnetic properties of the surface of submarine basalts are therefore largely a function of cooling history, rather than any upper mantle phenomenon. The new data confirm that deuteric or post-cooling alteration of basalts and ultrabasic rocks can be magnetically destructive: chloritization is always associated with a decreasing intensity of magnetization and Q ratio. Spilitization is similarly destructive. The magnetic effect of serpentinization, however, is not uniquely predictable. The magnetic data for submarine ultrabasic rocks show much variation, but are too limited for further generalization.