High density and high confinement operation in ELMy H-mode is confirmed at or above the normalized parameters foreseen for the ITER operating point (H 98(y,2) ∼ 1, n/n GW ∼ 1, β N > 1.8 at q 95 ∼ 3). The scaling of the ELMy H-mode with β N could be more favourable than that predicted by the IPB98(y,2) scaling. In ELMy H-mode, ion cyclotron current drive (ICCD) control of large sawteeth stabilized by fast particle has been demonstrated and the underlying neo-classical tearing modes (NTMs) and sawtooth physics is being refined. At high-density, Type I ELMy H-modes show trends that would lead to marginally acceptable ELMs on ITER. Type II ELM regime has been produced, though under very restrictive conditions. Type III ELMy operation with radiation fractions up to 95% has been demonstrated by seeding of N 2 in H-modes and could extrapolate to Q = 10 ITER operation, albeit at high current (17 MA). The mitigation of Type I ELMs, nevertheless, remains a challenge. Considerable progress has been obtained in internal transport barrier (ITB) plasmas, with operation at central densities close to the Greenwald density or/and low toroidal rotation or/and high triangularity. Demonstrations of full current drive and successful simultaneous real time control of safety factor and temperature profiles have been achieved in ITB plasmas. Physics of resistive wall modes (RWMs) has been compared with theory, showing favourable scaling for ITER. High β N ∼ 2.8 operation of hybrid modes (also called improved H-modes) has been obtained with dominant neutral beam heating. Hybrid modes with dominant ion cyclotron resonance heating (ICRH) have also been achieved.Trace tritium experiments yielded valuable information on particle transport in H-mode, ITB and hybrid regimes. In Type I ELMy plasmas, successful tests of the conjugate-T ICRH scheme have been achieved as well as lower hybrid coupling at ITER-relevant 10-11 cm distances. Reduced D and T fuel retention has been observed, which could relate to operation with vertical targets in the divertor and/or lower (ITER-like) vessel temperature. It is confirmed that erosion occurs predominantly on the main chamber surfaces, with possible benefits for T retention in ITER, although consequences for the metallic first wall lifetime need to be assessed. Disruption and ELM studies indicate that transient power deposition could be less constraining than expected for the ITER divertor, but more challenging for the metallic first wall. Alpha particle tomography and direct observation of alpha particle slowing down have been made possible by γ -spectroscopy. Measurements of Alfvén cascades have been improved by a new interferometric technique. Promising tests of ITER relevant neutron counting detectors have been conducted.