The Oman ophiolite has an axial volcanic suite and associated sheeted dike complex similar in composition to modern mid-oceanic ridge basalt (MORB). Its internal structure is regarded by many as being directly comparable to ocean lithosphere from the East Pacific Rise. However, there has long been controversy over the geodynamic setting in which the ophiolite formed, and the extent to which the analogy can be drawn, because the MORB-like axial volcanics are overlain by lavas that include depleted arc tholeiites and boninites. To some, this implies that the entire ophiolite formed above a subduction zone; others maintain that it formed at a true open-ocean MOR, and that the water required to generate the arc-like magmas derived from descending near-axis hydrothermal fluids or from ancient subduction. A popular compromise posits that the axial suite formed at a true MOR and the later magmatism documents the initiation of obduction. We test these models by reexamining the “MORB-like” character of the early axial lavas and dikes. We show that fractionation trends require the presence of water at concentrations significantly higher than any open-ocean MORB; instead, trends are identical to those of backarc basin and intraoceanic forearc volcanics. By showing that the entire ophiolite formed in a hydrous system, we rule out all models in which the Oman ophiolite was generated at an open-ocean MOR; instead, it formed at a submarine spreading center above a (probably newly initiated) subduction zone.