If sufficient tritium cannot be produced and processed in tokamak blankets then at least two alternatives are possible. Tritium can be purchased; or reactors with reduced tritium (RT) content in the plasma can be designed. The latter choice may require development of magnet technology etc., but we show that the impact on the cost-of-electricity may be mild. Cost tradeoffs are compared to the market value of tritium. Adequate tritium production in fusion blankets is preferred, but we show there is some flexibility in the deployment of fusion if this is not possible.