The discovery of massive sulfide ore deposits at certain sites at active seafloor spreading centers provides a basis for considering the energetics of processes that have concentrated similar large ore deposits that have been preserved in volcanic rocks on land. In this paper we construct transport models based on hydrothermal convection systems in the oceanic crust. We investigate models in which the circulation is driven by heat extracted from the permeable crustal rocks themselves as well as by heat conducted through the roof of a replenished, vigorously convecting crustal magma chamber. If the hydrothermal fluids carry 100 ppm dissolved iron as found in black smokers on the East Pacific Rise and exist at 350 ¿C or greater, calculations show that the heat content of the oceanic crust is insufficient to account for ore deposits of 3 Mt or greater. Heat extraction through the roof of a magma chamber, in conjunction with the formation of a thin (~1 km) layer of plated gabbro, may account for ore bodies of the order of 3 Mt, provided the permeability of the oceanic crust in the magma-hydrothermal fluid contact zone is 10-15 m2 or less. If the conducting boundary layer were only 1 m thick, locally, for 103 years, the permeability could be as high as 10-13 m2. The permeability in the discharge zone could, conceivably, be two orders of magnitude greater than the permeability at depth. More work is needed on the details of the effects of magma chamber replenishment and off-axis dike injection and on the role of magmatic fluids. Very large ore deposits (i.e., ~100 Mt) are probably formed in multiple hydrothermal cycles.