In the described research project the fatigue monitoring technique is applied on three 500kW wind turbines of the same type operating under different external conditions, i.e. stand alone, wind farm and complex terrain conditions. Respective sites have been selected in Germany (stand alone and wind farm) and in Greece (complex terrain). During the measurement campaigns three new databases have been created holding on-line monitoring data sets and time series data of the wind turbines' key loads. As a fourth data source, time series measurements from the wind farm at Alsvik, Sweden, have been used for the project work. The central aspect of the work has been to establish "footprints" of the measured load quantities for varying external conditions and to develop and accumulate skills and experience in reading the load information stored in such fatigue load "footprints". The term "footprint" refers to the rainflow cycle frequency spectra of the observed load quantity recorded during a representative time interval together with a set of parameters describing the external and operational conditions during that time interval. In fatigue monitoring the rainflow counting data reduction technique is applied to the measured load samples on-line, reducing hardware memory and off-line evaluation demands. It has been attempted to introduce a framework of a few statistic parameters that describe the fatigue load footprint and also relate to external physical conditions (average wind speed, turbulence etc.). In addition to the traditional formulations of statistic parameters in terms of time series statistics, special parameters adapted to on-line rainflow counted data sets have been examined. On-line fatigue footprint monitoring has so far been applied as a diagnostic tool. In the project the development of a scheme has been started that shall enable researchers to normalise the footprinting results and furthermore to extrapolate them to external conditions other than those present during the measurement campaign. A second focus of the work has been to incorporate the findings into a refined monitoring method and to exploit this method for assessment of wind turbine fatigue in wind farm and complex terrain operation. After completion of the research project the methodology has been applied in commercial services to assess the fatigue loading on individual components, such as the wind turbine main shaft. The method is well suited to supply relevant fatigue load data to the wind turbine designer at a comparably low cost.