Free running manoeuvring model tests are one of most used model tests to quickly assess the manoeuvrability of a ship. The results of these tests may be subject to scale effects: the fact that the model is smaller than the real ship, causes a number of dissimilarities (apart from the similarities). One of the known scale effects is the 'too high frictional resistance' at model scale. This causes with free running tests that the propeller is working too hard and that consequently the flow over the rudder is higher than it should be. At the same time, the wake fraction on model scale is higher due to a too thick boundary layer. This causes a lower velocity at the location of the rudder, which counteract (partly) the too high velocity at the location of the rudder due to the propeller thrust. There are several measures that can be taken to counteract this phenomenon, one of them recently being proposed by Ueno et al. (2013), and demonstrated by tests. The present paper reports an investigation on the magnitude of the scale effect. Through a systematic parametric study, the effect of the scale on overshoot angles and tactical diameter is quantified.