A new dataset of nearly 15,000 oceanic heat flow measurements is analyzed to determine the conductive heat loss through the seafloor. Many heat flow values in seafloor younger than 60 Ma are lower than predicted by models of conductively cooled lithosphere. This heat flow deficit is caused by ventilated hydrothermal circulation recharge and discharge at crustal outcrops or areas of thin sedimentary cover. Filtering heat flow data, retaining only sites with > 400 m of sediment cover and located > 60 km from the nearest seamount, minimizes the effect of hydrothermal ventilation. Filtered heat flow exhibits a much higher correlation coefficient with seafloor age (up to 0.95 for filtered data in contrast to 0.5 for unfiltered data) and lower variability (reduction by 30%) within an age bin. A small heat flow deficit still persists at ages < 25 Ma, possibly as a result of global filtering limitations and incomplete thermal rebound following sediment burial. Detailed heat flow surveys co-located with seismic data can identify environments favoring conductive heat flow; heat flow collected in these environments is higher than that determined by the filtered global dataset, and is more consistent with conductive cooling of the lithosphere. The new filtered data analysis and the growing number of site specific surveys both support estimates of global heat loss in the range 40–47 TW. Display omitted ► We analyze ~ 15,000 ocean heat flow observations globally. ► Sediment and seamount filters improve conductive heat loss estimates. ► Site-specific analyses provide robust estimates of conductive heat loss. ► Filtered heat flow vs. age is consistent with conductive plate cooling models. ► Estimated heat flow vs. age is consistent with oceanic heat loss, 40–47 TW.