Rechargeable lithium-air (O2) batteries are receiving intense interest because their high theoretical specific energy exceeds that of lithium-ion batteries. If the Li-O2 battery is ever to succeed, highly reversible formation/decomposition of Li2O2 must take place at the cathode on cycling. However, carbon, used ubiquitously as the basis of the cathode, decomposes during Li2O2 oxidation on charge and actively promotes electrolyte decomposition on cycling. Replacing carbon with a nanoporous gold cathode, when in contact with a dimethyl sulphoxide-based electrolyte, does seem to demonstrate better stability. However, nanoporous gold is not a suitable cathode; its high mass destroys the key advantage of Li-O2 over Li ion (specific energy), it is too expensive and too difficult to fabricate. Identifying a suitable cathode material for the Li-O2 cell is one of the greatest challenges at present. Here we show that a TiC-based cathode reduces greatly side reactions (arising from the electrolyte and electrode degradation) compared with carbon and exhibits better reversible formation/decomposition of Li2O2 even than nanoporous gold (>98% capacity retention after 100 cycles, compared with 95% for nanoporous gold); it is also four times lighter, of lower cost and easier to fabricate. The stability may originate from the presence of TiO2 (along with some TiOC) on the surface of TiC. In contrast to carbon or nanoporous gold, TiC seems to represent a more viable, stable, cathode for aprotic Li-O2 cells.