Abstract In many places heat flow decreases towards the edges of young submarine sedimentary basins, and toward basement outcrops within them. Perhaps this fact contains some information on the thickness and permeability of the convective zone in the ocean crust. A laboratory experiment was conducted by covering 88% of an 8 m 2 porous bed with an insulating layer made up of polymethylmethacrylate slabs. Heat flow through these slabs was obtained by pushing a thermistor-tipped snake between them and the bead-and-water porous bed beneath. At all convective fluxes studied from 24 to 2987 W, heat-flow depression only extended one-half of the porous-bed thickness into the insulated zone. The proportion of the total heat flux passing through the insulated area declined from 62% near the onset of convection to 27% at the highest fluxes attained. However, the flux transported convectively to the base of the insulating layer remains close to the value expected from the temperature difference, in spite of the larger flux delivered to the thermally-exposed zone nearby. The ability of the experimental apparatus to deliver the two separate fluxes is due to its highly-conductive, constant-temperature base: enhanced heat transfer due to cold flows from the thermally-exposed area leaves enough physical space for the modest plume flux needed by the resistive cover. Were the bottom boundary resistive, as in nature, the high heat flux from the exposed area would depress the surface-temperature of the base and thus the flux available for the insulated zone. There appears to be no mechanism in the porous flow itself that causes heat flow depression away from the edge of the resistive zone, and thus the marine heat-flow data contain little information about the thickness of the permeable zone in the oceanic crust.