Four stratigraphic (vertical) transects were obtained from the gabbroic and ultramafic section of the Samail ophiolite (middle Cretaceous age). The transect, through Wadi Kadir was the most detailed; 65 independent oriented cores were obtained from 22 sites on a 14-km traverse. From field mapping we infer that this traverse corresponds to a vertical section of 6.0 km, which includes sheeted dikes, plagiogranites, isotropic high-level gabbro, transitional gabbro, cumulate gabbros, cumulate ultramafics, and ultramafic rocks. There are some noticeable variations in magnetic mineralogy with depth: (1) primary exsolved magnetite is found mainly in the late stage high-level gabbro, and (2) secondary magnetite is found mainly in the upper 3.5 km of the cumulate gabbros and in the cumulate ultramafics. Most of the gabbros have a stable magnetization direction that is almost horizontal and slightly west of north. The median natural remanent magnetization (NRM) intensity for the cumulate gabbros is 0.93 A/m. The median Qn value is 3.0 with most values falling in the range of 1 to 5. The ultramafic rocks below the gabbroic layer are more strongly magnetized due to higher concentrations of secondary magnetite, and their magnetization is more stable to AF demagnetization than that of the overlying gabbros. It is not clear what fraction of their magnetization was acquired beneath the seafloor because some of the serpentinization is thought to be of subaerial origin. Hysteresis measurements on the gabbros suggest that the magnetic carriers in the gabbros are a few tenths of microns in diameter (pseudo single domain). The magnetic properties of the cumulate gabbro section divide it into an upper intensely magnetized layer (3.5 km thick) and a lower weak layer (2 km thick). This was also found for the Point Sal ophiolite (California). The low intensities below 2 km are in agreement with petrological observations of a minimum in the concentration of secondary magnetite. The upper intense layer can contribute over one half of the amplitude of any marine magnetic anomaly which might have existed when the ophiolite was submarine. The serpentinized rocks also are possible contributors to anomaly amplitude. Consideration of the probable variations in the magmatic and hydrothermal history of the ocean crust and comparison of our results to other studies indicate that it may not be possible to specify a generalized magnetic source model for the plutonic section of the ocean crust.