INTERPRETATIONS of the observed electrical conductivity of the upper mantle have relied on laboratory studies of dry olivine and basaltic melts1,2. These studies indicate that the conductivity of the asthenosphere is higher than that of olivine, suggesting that partial melting or other high-conductivity mechanisms such as grain-boundary phases are required2,3. Recent experimental studies4–6, however, have demonstrated that various kinetic processes in silicate minerals are greatly enhanced by the dissolution of hydrogen. As hydrogen is probably present in the upper mantle, studies on dry olivine may not therefore be relevant to the real Earth. Here I use experimental data on the solubility and diff usivity of hydrogen in olivine5,7 to estimate its effect on the electrical conductivity. I conclude that the observed high conductivity in the asthenosphere can be attributed to solid-state conductivity in olivine if a small amount of hydrogen is present. Partial melting of the asthenosphere on a global scale is therefore not required.