Neutronics experiments are performed at JET for validating in a real fusion environment the neutronics codes and nuclear data applied in ITER nuclear analyses. In particular, the neutron fluence through the penetrations of the JET torus hall is measured and compared with calculations to assess the capability of state-of-art numerical tools to correctly predict the radiation streaming in the ITER biological shield penetrations up to large distances from the neutron source, in large and complex geometries. Neutron streaming experiments started in 2012 when several hundreds of very sensitive thermo-luminescence detectors (TLDs), enriched to different levels in (LiF)-Li-6/(LiF)-Li-7, were used to measure the neutron and gamma dose separately. Lessons learnt from this first experiment led to significant improvements in the experimental arrangements to reduce the effects due to directional neutron source and self-shielding of TLDs. Here we report the results of measurements performed during the 2013-2014 JET campaign. Data from new positions, at further locations in the South West labyrinth and down to the Torus Hall basement through the air duct chimney, were obtained up to about a 40m distance from the plasma neutron source. In order to avoid interference between TLDs due to self-shielding effects, only TLDs containing natural Lithium and 99.97% Li-7 were used. All TLDs were located in the centre of large polyethylene (PE) moderators, with Li-nat and Li-7 crystals evenly arranged within two PE containers, one in horizontal and the other in vertical orientation, to investigate the shadowing effect in the directional neutron field. All TLDs were calibrated in the quantities of air kerma and neutron fluence. This improved experimental arrangement led to reduced statistical spread in the experimental data. The Monte Carlo N-Particle (MCNP) code was used to calculate the air kerma due to neutrons and the neutron fluence at detector positions, using a JET model validated up to the magnetic limbs. JET biological shield and penetrations, the PE moderators and TLDs were modelled in detail. Different tallying methods were used in the calculations, which are routinely used in ITER nuclear analyses: the mesh tally and the track length estimator with multiple steps calculations using the surface source write/read capability available in MCNP. In both cases, the calculated neutron fluence (C) was compared to the measured fluence (E) and hence C/E comparisons have been obtained and are discussed. These results provide a validation of neutronics numerical tools, codes and nuclear data, used for ITER design.