Calculations were performed to investigate how the thermonuclear energy gain for inertially confined fusion varies as a function of the tritium inventory. Results showed that for large ρR pellets (25 g/cm2) the yield was relatively insensitive to the D: T ratio, so that a reduction by a factor of five in the amount of tritium produced only a 15% reduction in yield. An attempt was made to further reduce the sensitivity by replacing some of the deuterium by 6Li. The idea was to breed tritium by neutron capture within the pellet and then to burn it in situ. However, the breeding rate was found to be too slow for this process to make any significant contribution to the burn. Lithium was not able to improve upon deuterium as a fuel.