The accuracy with which seafloor heat flow is determined depends on the temporal stability of bottom water temperature. Indirect tests for stability are provided most commonly by observing the uniformity of heat flow with depth. This criterion is met to a high degree of certainty at two sites in the eastern North Pacific Ocean, where colocated high-quality probe and borehole heat flow data can be compared. A more direct test for stability is provided by long-term observations of bottom water temperature. Previously published records and new data show temperature variations of only a few hundredths of a degree at sites in the central and eastern North Atlantic and the eastern North Pacific. Resultant gradient perturbations are geothermally insignificant (<5 mK m⁻¹) at depths greater than 1–2 m below the seafloor, consistent with the uniformity of heat flow with depth observed in these areas. Geothermally problematic bottom water temperature variations are observed or inferred in the western North and South Atlantic and western South Pacific. Variations range up to ±0.15 K and are capable of producing gradient perturbations of up to 5 mK m⁻¹ at depths as great as 5 m below the seafloor. While these data are instructive, their distribution is not adequate to provide general guidelines for estimating geothermal gradient perturbations. Data from shallower sites are needed in all oceans to define depth limits of acceptable bottom water temperature variability, and from other deep-ocean locations where near-source bottom water transients or vigorous deep-water circulation dynamics are likely to be present.