Various types of models can be used in ship manoeuvring simulation in harbours and fairways. The most important distinction is physical (scale) models and mathematical (numerical) models. Physical models can be used in two ways, viz. (1) steered by a human being, either on board of the ship or via a camera on board and remote control by the pilot on land, and (2) by using an autopilot system, where the ship follows a predetermined path and the autopilot system generates the rudder and engine orders to the ship. Mathematical models can be incorporated in a ship manoeuvring simulator where real pilot/helmsman's orders are entered into the model and where the modelled ship reacts to these orders. Similarly to physical models with an autopilot system, mathematical models can also be provided with an autopilot system; in this case also the ship has to follow a predetermined line, and the autopilot generates orders to the ship in order to follow this line. The difference between runs made in man-steered models on the one hand, and autopilot models, on the other, is that runs made with autopilots are, in principle, completely repeatable in the same conditions whereas man-steered models contain an important stochastic element and are therefore, in principle, unrepeatable. An important point in physical scale models is the fact that they are not only subjected to length scale, but, due to this, also to time scale (being the root of the length scale for Froude models). This means that steering of a physical model by a human being is subject to an important time-scale factor (the word "important" is used as in most cases this time scale will be 5 or more, given a length scale of 25 or more). This means an important disadvantage of using such models in simulation processes. Physical models with autopilot systems do not suffer from this disadvantage. It should be borne in mind, however, that the use of autopilot systems gives limited information only and a physical model is considered quite costly for such limited information. The advantage of physical models is that there is less doubt, compared to numerical models, about the validity of the model. Physical models are, therefore, still useful in situations where insufficient mathematical knowledge is available, or where physical processes can be modelled only by very complicated and extensive mathematical descriptions, making the mathematical model slow and sometimes unreliable. Examples of this are complicated bank suction situations and the passing or taking-over of other vessels with strong interaction effects. Physical ship manoeuvring models can sometimes be combined with hydraulic models in harbour design (current models, wave penetration models), and then their use can be more obvious. As numerical models become more and more common also in hydraulics, the combination of a physical hydraulic model and a physical ship manoeuvring model cannot very often be realised anymore. Mathematical models are nowadays quite common in ship manoeuvring. They are used in ship manoeuvring simulators and fast-time mathematical models. The most important advantage of a ship manoeuvring simulator over a scalemodel is that it allows real-time simulation, not only since the problem of time-scale is avoided, but also since it allows the representation of the ship, the bridge and the environment on real-world scale. The mathematical models describing the physical process of ship manoeuvring are, depending on the type and extent of the model, considered adequate for not too complicated situatiohs. Most mathematical models are based on physical model tests to determine the various coefficients in the model. The model which is most used is the Abkowitz model; the number of coefficients in this model is not fixed. Due to this, there will be a difference in accuracy between various models. The importance of this depends on the problem which is modelled. Equality between numerical models and prototypes can be shown only by comparison of results in well-known and not too complicated environmental conditions. The latter is mostly only true in deep water, and the comparison is limited to turning-circle, zig-zag and spiral tests and acceleration and deceleration tests. Most mathematical models are able to present proper comparisons in this respect. The behaviour of ships in shallow water is never checked in the real world due to the risks. Therefore, numerical models have to be checked. in such condition by comparison with scale model tests. These comparisons give sufficient evidence of the reliability of numerical models. The quality of ship manoeuvring simulation models is not limited to the quality of the mathematical model describing the ship's motions and path alone. An important point is also the modelling of controls, like rudder, engine and tugs. This modelling is not limited to physics alone, but also to adequate modelling of the real behaviour of rudder, engine, propeller and tugs in terms of reaction characteristics. For ship manoeuvring simulators, there is also the similarity between simulator outfit and the real ship. This has regard to the presence of a bridge, the quality of the outside view, the presence of sound and possible vibrations on the bridge, short-periodic motions due to waves, communication aspects, etc. Most important in this respect are the presence of a real ship's bridge and the outside view. The relevant aspects in the outside view are: a sufficiently large view angle, a realistic view regarding motions (especially rate of turn), a realistic quality of the view with not too much details, and a picture which is as good or as bad as reality. Attention in ship manoeuvring simulation studies is generally focussed on the validity of the mathematical ship manoeuvring model. Although this is a very important aspect, other aspects deserve also attention. This applies to a proper problem formulation, the experimental design method, the choice of subjects (pilots and/or masters), and the method of data analysis and drawing conclusions from the investigation. Moreover, there is still the aspect that it is often not necessary to have a ship in the simulator, which is exactly similar to one specific prototype vessel. In reality ships differ also and pilots are used to this. There are many different simulator designs varying from simple micro-simulators to very sophisticated simulators. It is the working group's conviction that all of them serve a certain purpose and can be used in one way or another in the design process of a harbour or fairway.