Hydrogen strongly influences mantle rheology and is therefore an important factor in the growth and stability of cratons. Hydrogen also strongly affects electrical conductivity so it is possible to infer the hydrogen content of the lithospheric mantle in-situ and test models of craton formation using magnetotelluric data. Tanzania is an ideal natural laboratory to test hypotheses on lithospheric hydrogen content since it contains regions with very different tectonic regimes including the stable Tanzanian Craton and the East Africa Rift that is reworking lithosphere previously deformed in the East African Orogeny. Additionally, the lithosphere is well sampled by voluminous xenoliths that constrain lithospheric composition and the geotherm, which also affect electrical conductivity. Hydrogen contents were calculated for two locations in Tanzania: the first in the stable central Tanzanian Craton and the second on the eastern margin of the craton where incipient rifting is occurring. The central Tanzanian Craton was found to have a high lithospheric mantle water content of ~10 -2 wt% which is comparable to that of the oceanic asthenosphere and is hard to reconcile with the long-term survival of the craton. It is possible that the water was introduced into the lithosphere recently by kimberlite volcanism or that, if the lithosphere has had a high water content throughout its history, the central craton has been shielded from deformation by weaker orogens that surround it. The eastern margin of the craton has a water content of 10 -3 to 10 -4 wt% throughout much of the lithospheric mantle that decreases to 10 -4 to 10 -5 wt% at the base of the lithosphere and at depths corresponding to the uppermost plume head. Xenolith data show evidence for partial melting of the plume head and the base of the lithosphere in this dehydrated region. The partial melting and dehydration of a plume head beneath a craton is a present-day observation of the processes that may have formed cratonic roots.