Particle retention is a major constraint for future fusion devices like ITER in which the amount of tritium will be strictly limited for safety reasons. In the EU Task Force on plasma wall interaction, efforts are underway to investigate the gas balance, the particle retention and removal in fusion devices. Gas balance in JET, ASDEX Upgrade, TEXTOR and Tore Supra are reported in this paper. In all these devices, a peak in the wall loading is s observed, at the beginning of the plasma, which is attributed to the saturation of the area in contact with the plasma. These particles are always recovered at the end of the plasma (dynamic retention) while for longer plasma operation, the particle retention can become proportional to the discharge duration. The effects on the particle retention by different fuelling methods, gas puffing, pellet injection and Neutral Beam Injection (NBI) are reported. ,there is only a very weak reduction in particle retention with pellets compared to gas injection, while with NBI a transient wall depletion is always observed but accompanied by a density drop which requires additional gas puffing to recover the target plasma density. For all the devices, the recycling flux dominates the particle fluxes and neither pumping nor fuelling allows to modify/control the recycling flux and consequently the retention flux. The particle recovery between pulses by gas release is always similar in the absence of disruptions. However, for longer plasma durations this contribution becomes negligible in the overall balance. Finally, conditioning methods (glow discharges) and discharges cleaning show a particle recovery which is independent of the particle retention in the previous plasma operations.