The subject of aerodynamics is of major importance for the successful deployment of wind energy. As a matter of fact there are two aerodynamic areas in the wind energy technology: Rotor aerodynamics and wind farm aerodynamics. The first subject considers the flow around the rotor and the second subject considers the (wake) flow within a wind farm. For both areas calculational models have been developed which are implemented i rotor design and wind farm design codes respectively. Accurate rotor design codes enable a reliable design of wind turbines and an optimization towards a higher energy production and lower loads, i.e. towards a lower cost of energy. They are also required to avoid design errors and hence to reduce investment risks of wind turbine manufacturers. Accurate wind farm design codes are needed to predict the production losses and the load increase on turbines in a farm due to wake effects. They also support the optimization of wind farms (e.g. through farm control) by which the energy losses and the load increase from wake effects (and consequently the costs/kWh) are minimized. For both areas the complexity of models range from engineering methods to very advanced Computational Fluid Dynamics (CFD) methods. The term engineering method is meant to indicate a model which casts a complicated flow phenomenon into a transparent form. This generally goes together with an economic computer usage. In this respect it is very important to realize that wind energy design calculations are inherently very time consuming by which advanced CFD models are still beyond the routine possibilities of industry. As such engineering methods form the only alternative for that purpose. The main aim of the present thesis is then to describe several developments of the last 25 years which have led to the present generation of aerodynamic engineering models. It will be shown that much progress has been made both on the field of rotor aerodynamics as well as on the field of wind farm aerodynamics and that this progress was highly supported by the fact that dedicated aerodynamic measurement data have become available. The progress is illustrated by the engineering models which are developed and validated by ECN in several large (inter)national cooperation projects in which these measurements played an important role. The author of this thesis was heavily involved in these projects and often acted as coordinator. Since these projects were performed in close cooperation with other institutes (which used different types of models), the activities of the author can be placed in a wider context. The first part of the thesis is devoted to rotor aerodynamics. Basically the subject of rotor aerodynamics can be subdivided in two parts: The first part deals with the global flow field around a wind turbine. This type of modelling is called induction aerodynamics, since its main goal is to determine the induced velocities at the blade. The second part deals with the loads on a wind turbine blade as a response to this flow situation and is called blade aerodynamics. Current engineering models for rotor aerodynamics topic are built around the Blade Element Momentum (BEM) theory. The Blade Element Momentum theory in itself is very basic, e.g. it is derived for 2-dimensional, stationary, homogenous and non-yawed conditions. For this reason several engineering models have been developed which overcome these simplifications and which act as add-on's to the basic BEM theory. These engineering add-on's have been developed for the field of blade aerodynamics and for the field of induction aerodynamics. In this thesis a comparison is made between current engineering models and the engineering models from 25 years ago. The engineering methods from 25 years ago were not much more than the very basic BEM theory with a Prandtl tip loss correction and a turbulent wake correction. Moreover a tower shadow model based on a dipole model and a 'geometric' correction for cone and tilt angle were included, while yaw was modelled with the advancing and retreating blade effect only. Since then the models for airfoil aerodynamics have been improved by adding unsteady and three-dimensional effects. These unsteady effects can be divided in viscous dynamic stall effects and non-viscous effects at low angles of attack. The three-dimensional effects occur at the inner part of the blade where stall is delayed and at the outer part where the tip decreases the loads. In terms of induction aerodynamics, models have been added for dynamic inflow, the azimuthal variation of the induced velocity at yaw and a model for root losses. The progress in the rotor aerodynamic engineering models from ECN is mainly described along results of four subsequent IEA Tasks: IEA Task 14 and 18, IEA Task 20 and IEA Task 29(Mexnext). An IEA Task (sometimes called an IEA Annex) is a cooperative project carried out under auspices of the International Energy Agency IEA. The goal of IEA Tasks 14 and 18 was to create a database of detailed aerodynamic measurements which all have been taken on turbines under atmospheric conditions. The goal of IEA Task 20 was to analyze the measurements which have been taken by the National Renewable Energy Laboratory NREL on a 10 meter diameter wind turbine which was placed in the very large NASA-Ames wind tunnel. Finally IEA Task 29(Mexnext) analyzed the measurements which have been taken in the EU Project Mexico on a wind turbine rotor with a diameter of 4.5 meters placed in the Large Low Speed Facility (LLF) of the German Dutch Wind Tunnel (DNW). In all of these experimental programs pressure distributions were measured at different locations along the rotor blades. Moreover the Mexico experiment mapped the flow field upstream, in and downstream of the rotor plane. The detailed aerodynamic measurements from the IEA Tasks were found to be very useful in the development, improvement and validation of these engineering models because they made it possible to extract aerodynamic phenomena which are hidden in the very global information from conventional measurement programs. It is concluded that only detailed aerodynamic measurements may be used for validation of aerodynamic design models: A validation on basis of global turbine(blade) loads does not give a decisive answer on the accuracy of aerodynamic models due to the fact that 'compensating errors' may occur. Moreover it will be shown that the measurements revealed several shortcomings in aerodynamic engineering methods which partly could be 'repaired', sometimes with the help of more refined models. Several recommendations are made on rotor aerodynamics. This includes some specific further improvements which are still possible to the current state of engineering models. Amongst other things, models for the annulus averaged induction at yaw, tip loss effects and time constants at dynamic inflow can be improved further. These improvements can be established by calibrating engineering methods to results from more advanced aerodynamic models (e.g. CFD or free vortex wake methods). The background for this recommendation lies in the fact that the validation of these advanced aerodynamic models with the detailed aerodynamic measurements from the IEA Tasks showed a clear added value from such methods on these fields. Moreover it is concluded that three-dimensional and unsteady effects on the drag deserve more attention. However the most important recommendation is related to the observation of an unbalance in the aerodynamic wind energy society: Much effort is spent on the development of aerodynamic models (often of little mutual differences) but the amount of experimental validation material is (too) limited. Therefore it is recommended to intensify the activities on rotor aerodynamic measurements in both the wind tunnel and the field. Special attention should be paid to the measurement of those phenomena which, until now, are still largely concealed (e.g. boundary layer phenomena) or unclear (e.g. the relation between blade loads and underlying flow field which is found puzzling in the Mexico experiment). The present thesis also describes the progress which has been made on the field of wind farm aerodynamics. Opposite to the situation for rotor aerodynamics, where the BEM model can be appointed as the main model, the variety of models for wind farm aerodynamics is much larger. This is partly due to the fact that a wind farm aerodynamic model should cover much more aspects: It should model both the aerodynamic behavior of the rotor (which generates the wake) as well as the turbulent wake downstream of this rotor. The fact that calculational time is such an extreme constraint adds to the diversity: As a consequence CFD modelling of wind farm aerodynamics often only refers to the modelling of the wake and not to the modelling of the rotor. It also makes that wind farm and rotor aerodynamics are sometimes considered to be fully separate subjects. This is seen as an undesired development since the aerodynamics of the wake is largely determined by the aerodynamics of the rotor standing in front of the wake. In this thesis the main characteristics of the wake flow behind a wind turbine are described together with a survey of wind farm aerodynamic models. Most of the attention is focussed on an intermediate between the very basic models and the CFD codes, i.e. the parabolized wake models. These models are relatively economic in computer usage (by which they are still considered to be engineering models) where they model the so-called far wake in a physically accurate way. The disadvantage lies in the fact that they generally need an empirical treatment of the near wake. This again goes together with a very simple modelling of the rotor. The progress in wind farm aerodynamic models is then illustrated with ECN's wind farm design code Farmflow (based on the former Wakefarm wake model) which combines a parabolized k-epsilon turbulence model for the far wake with results from a physical free vortex wake method for the near wake. The measurements on wind farm aerodynamics used in this thesis mainly come from the ECN Wind Turbine Test Site Wieringermeer, EWTW. This research farm consists of five wind turbines in a line set up with a rated power of 2.5 MW and a rotor diameter and hub height of 80 meter. The turbines are extensively instrumented, where a meteorological mast is available to measure the free stream or the wake conditions. A major advantage of these measurements lies in the research environment by which data have been recorded over a very long period of high quality. The EWTW measurements revealed various new wake aerodynamic phenomena and they offered validation material for the improvement and validation of the Farmflow code. The observations on the EWTW farm are compared with those on large off-shore wind farms, the measurements of which were supplied within the EU project Upwind. In the EWTW line set-up the largest power loss due to wake effects (and hence the lowest overall power) appears at the second turbine in the farm. The turbines deeper in the farm have a slightly higher power. This is opposite to the situation in large off-shore wind farms where the power keeps decreasing for turbines deeper into the farm. This can be explained by lateral wake effects and the size of those large (array) wind farms. The power behavior in both the EWTW as well as in the large array wind farms was predicted well with Farmflow. Several conclusions on wind farm aerodynamics are drawn. The most important conclusion is that as for the situation on rotor aerodynamics, much progress has been achieved over the past decades. This is illustrated with the developments from Wakefarm to Farmflow. In the beginning of the 1990's only single wakes were considered. These were modelled with a very simple approach: The wind turbine was represented by an actuator disc with a near wake model based on momentum theory (and later empiricism). The far wake was modelled with a turbulence model tuned for non wind energy applications. Since then the near wake models has been refined and multiple wake effects are taken into account in both axial and lateral direction. Furthermore the turbulence model has been calibrated for wind turbine wake situations. For the development of wind farm engineering models in general it is very important that some CFD models entered the (research) scene in which the rotor is modelled with more advanced methods than the actuator disc approach (e.g. with actuator lines). Such advanced models can now be used for calibration of more simple models. Several subjects for wind farm aerodynamics have been identified which still need more attention. As such it is recommended to intensify research on these fields. This holds amongst other things for the validation and improvement of multiple wake models and near wake models in multiple wake situations. Also the interaction of wind farms with the outer atmosphere deserves more attention. Moreover there is a need to refine the turbulence models for wind farm aerodynamics. Another main question to be answered is the importance of rotor aerodynamics for wake aerodynamics. More specifically it should be determined whether it is justified to model the rotor as an actuator disc. The answer to this question can be found by comparing results from CFD codes, which models both the rotor and the wake in a detailed way, with results from a similar code in which the rotor is replaced by an actuator disc. As for the situation on rotor aerodynamics it is again concluded that progress on the field of wind farm aerodynamics is hampered by a shortage of high quality validation material. For this reason it is recommended to intensify the measurement activities for wind farm aerodynamics. In this thesis minimum requirements for such measurement programs are given. Measurements anyhow need to be done on full scale wind farms, preferably in combination with wind tunnel measurements. The first type of measurements yield representative information but generally lack a sufficient degree of detail for a complete interpretation of the wind farm aerodynamic problem. Furthermore field measurements are difficult to interpret due to the stochastic turbulent environment in the free atmosphere. The second type of measurements can yield very detailed and easy interpretable information but the scale of the model turbines is far too small. An interesting intermediate is then the so-called ECN scaled wind farm. This farm consists of 10 wind turbines with a rotor diameter of 7.6 m and a rated power of 10 kW. The farm is heavily instrumented where the size is sufficiently large to make the results at least to some extent, representative for full scale situations. The combination of full scale measurements, scaled farm measurements and wind tunnel measurements then forms the most complete experimental base for wind farm aerodynamics even though each type of measurements has its own drawbacks.