We present 206 new heat flow measurements in the Indian Ocean. These and approximately 300 previously published heat flow values are individually evaluated for sedimentary environment and instrumental performance . The relationship between average heat flow and age is found to be little affected by selection of the most reliable experiments, although the scatter about the mean is significantly lowered. The variation of mean heat flow with age is found to be very similar to that in the eastern Pacific and Atlantic oceans: There is a crestal low heat flow zone with large variability, a transition zone within which the heat flow increases from values considerably below to values in agreement with predictions from thermal models of the oceanic lithosphere, and a region where heat flow values are in accord with theoretical predictions. However, the transition zone occurs over different crustal ages from ocean to ocean: 40--60 m.y. in the Indian Ocean, 4--6 m.y. in the Gal¿pagos spreading center. 10--15 m.y. on the East Pacific Rise, and 50--70 m.y. on the Mid-Atlantic Ridge. The transition zone generally corresponds to a sea floor age where (1) sedimentary thickness increases to ?300 m, (2) sea floor roughness is significantly smoothed by sediment blanketing, and (3) the carbonate content of surface sediments decreases to ?40%. The transition zone occurs where water circulation in the oceanic crust stops affecting the surface heat flow strongly. There are two possible explanations for the transition. First, a change in composition from carbonate to siliceous sediments results in a decrease in bulk permeability. This combined with general thickening of the sedimentary blanket with aging results in the deposition of an impermeable layer which prevents the convective exchange of heat from the oceanic crust to the ocean. Second, hydrothermal flow within the oceanic crust is plugged by filling of circulation cracks in the oceanic crust. The fact that in several basins of the Indian Ocean the heat flow transition corresponds with the carbonate-siliceous boundary is support for the former mechanism. However, the fact that locations of increases in velocity of seismic layer 2A generally correspond to the transition regions in the Atlantic and Pacific oceans provides support for the latter mechanism. Heat flow measurements in the world's oceans allow us to calculate the variations of bulk permeability and basal temperature in the oceanic crust as a function of age and to evaluate the geochemical implications of the variation in these parameters between oceans. The combination of conductive heat flow and elevation versus age observations in old lithosphere demonstrates the deviation from t1/2 cooling in the Indian Ocean and indicates that the Mozambique and western Somali basins are considerably older than preliminary deep-sea drilling results suggest.