A transport code (TRANSP) is used to simulate future deuterium-tritium (DT) experiments in TFTR. The simulations are derived from 14 TFTR DD discharges, and the modelling of one supershot is discussed in detail to indicate the degree of accuracy of the TRANSP modelling. Fusion energy yields and 01 particle parameters are calculated, including profiles of the 01 slowing down time, the 01 average energy, and the AlfvBn speed and frequency. Two types of simulation are discussed. The main emphasis is on the DT equivalent, where an equal mix of D and T is substituted for the D in the initial target plasma, and for the Do in the neutral beam injection, but the other measured beam and plasma parameters are unchanged. This simulation does not assume that 01 heating will enhance the plasma parameters or that confinement will increase with the addition of tritium. The maximum relative fusion yield calculated for these simulations is QDT-0.3, and the maximum a contribution to the central toroidal 0 is PJO)-0.5%. The stability of toroidicity induced Alfvkn eigenmodes (TAE) and kinetic ballooning modes (KBM) is discussed. The TAE mode is predicted to become unstable for some of the simulations, particularly after the termination of neutral beam injection. In the second type of simulation, empirical supershot scaling relations are used to project the performance at the maximum expected beam power. The MHD stability of the simulations is discussed.