ITER steady state operation relies on Internal Transport Barrier (ITB) at a large radius (r/a>0.5) in order to improve the confinement of the standard scenario, by about 1.5 (H98y,2~1.5), allowing the total current to be reduced and making it easier to operate fully non-inductively. A minimum ~50% of bootstrap current (Ibs) has to be obtained, given the predicted external power available for Non-Inductive (NI) current drive. A high normalised beta, bN~3, and high density, ne~0.8·nGW, should also be achieved to preserve good performance. The location of the NI currents must be compatible with the q profile favourable for the ITB. Experiments have been carried out at JET to develop such plasmas in an ITER-like configuration, namely with q95~5 and high triangularity, d~0.4. On a target plasma with BT = 2.3T, Ip = 1.5MA, bN peak values up to 3, and ~2.7 if averaged over the ITB duration, have been reached to be compared with 4·li~2.8, with H98(y,2)~1 and ne~0.8·nGW. Preliminary TRANSP modelling indicate that the fraction of IBS is ~30-40%, with ~15-20% of NBI current. The best performances are obtained when both e- and i+ ITBs are present and are of very similar width rITB,e~rITB,i~3.6m (r/a~0.45) and strength, in terms of the normalized temperature gradients r*Te~r*Ti~0.021. q profile with negative magnetic shear is formed using Lower Hybrid Current Drive (LHCD) early in the discharge, and high NBI and ICRH power are added when qmin is just above 2, to trigger an ITB when qmin = 2. Puffing Ne+D2 mitigates large ELMs that are not compatible with a wide ITB. Nevertheless strong ELMs periodically degrade the energy content by more than 10% at the peak values of bN, limiting the performance. The ITB is maintained longer than 17 energy confinement times until detrimental MHD activity, linked to the loss of the optimum q profile, appears. This indicates that more NI current is required to slow down the q profile evolution. The effectiveness of the LHCD could be increased by reducing the edge density, which is higher in this regime than in the low delta configurations used for many previous JET ITB experiments. This could be obtained by optimizing the D2 and Ne puffing and by moving the plasma away from the walls, in order to reduce the particle recycling and refuelling.