An expanding spread profile (ESP) over 80 Ma Pacific Ocean crust near Hawaii was modeled with a one-dimensional solution for travel times, relative amplitudes, and sometimes frequency content. The resulting profiles of compressional and shear wave velocities and Poisson's ratio are similar to those determined by other ocean crustal studies except that the shear wave Q in the sediments is very low (10) and some unusual high-frequency, high-phase-velocity arrivals require a laminated structure within the mid-crustal section at about 3 km below the base of the sediment. The laminations are probably alternating layers of gabbro and ultramafic cumulates. The cumulates may be fractionation products from an upper level magma chamber that was present at the time of crustal formation. Reflections which may correspond to these laminations have been observed in reflection records both near this ESP and in the North Atlantic. Comparison of different ophiolite models for ocean crust reveal that the most commonly used velocity-lithology model is probably in error for the pillow and dike basalts, resulting in an interpretation of overly thick layers in previous studies of refraction lines. The preferred interpretation is that the average total thickness of the pillow and dike sequences is between 1.0 and 1.5 km. ¿1991 American Geophysical Union