Plasma rotation can have beneficial effects on tokamak fusion plasmas. For instance, plasma rotation can enhance the stabilising effect of a resistive wall. Recent experimental results suggest that the required speeds could be relatively modest. In contrast to in present day devices, heating by Neutral Beam Injection (NBI) is not expected to induce strong plasma rotation in ITER and future reactors. It is therefore of interest to consider other mechanisms that can give rise to plasma rotation. Intriguing observations of plasma rotation have been made in ICRF heated plasmas with little or no external momentum input, often referred to as intrinsic rotation. The mechanism behind the rotation is not yet well understood. Recent work aiming at developing a multi machine database indicates that the rotation velocity extrapolated to ITER conditions could provide a useful contribution to restive wall mode stabilization. However, there are large uncertainties, the dependence on many important factors (fast particles, ripple etc) is not clear. It is therefore of interest to investigate in more detail the origin of the intrinsic rotation.