TFTR experiments have emphasized the optimization of high performance plasmas as well as studies of transport in high temperature plasmas. The recent installation of carbon composite tiles on the main bumper limiter has allowed operation with up to 32 MW of neutral beam injection without degradation of plasma performance by large bursts of carbon impurities ('carbon blooms'). Plasma parameters have been extended to Ti(0) approximately 35 keV, Te(0) approximately 1.2 × 1020 m-3, producing D-D reaction rates of 8.8 × 1016 reactions per second. The fusion parameter ne(0)τETi(0) in supershot plasmas is an increasing function of heating power up to an MHD stability limit, reaching values of approximately 4.4 × 1020 m-3 sec keV. Peaked-density-profile hot-ion plasmas with the edge characteristics of the H-mode have been produced in a circular cross-section limiter configuration with ne(0)τETi(0) values characteristic of supershots, namely up to four times those projected for standard H-modes with broad density profiles. Reduced transport is also observed in the core of high-density ICRF-heated plasmas when the density profile is peaked. At the highest performance, the central plasma pressure in TFTR reaches reactor level values of 6.5 atmospheres. In these regimes, MHD instabilities with m/n = 1/1, 2/1, 3/2 and 4/3 are often observed concurrent with a degradation in performance. High βp plasmas with εβp ≈ 1.6 and β/(I/aB) ≈ 4.7 (%mT/MA) have demonstrated confinement enhancement over the low-mode confinement time with τE/τL approximately 3.5 and a bootstrap current of about 65% of the total plasma current. The best TFTR supershots in deuterium plasmas have QDD = 1.9 × 10-3, corresponding to an equivalent QDT approximately 0.31 if the plasma and beam parameters are exactly the same in a 50/50 D/T plasma. Performance enhancements in D-T due to higher heating power, higher density, lower Zeff, ion mass effect and optimum programming of the D-T mix in the beam extrapolate to D-T plasma performance near the breakeven regime, QDT* approximately 0.5 - 0.7, that would produce approximately 15 - 25 MW of fusion power. Lower Q (approximately 0.3) plasmas at higher temperatures are expected to produce alpha particle betas suitable for testing collective alpha instability theories.