The worldwide struggle to renew the human energy supply to high environmental friendly standards has led to creative and innovative solutions for the supply of `green' energy. The financial benefits of conventional energy sources restrain the widespread use of renewable energy sources, like wind or solar energy. This competition requires ingenuity and sometimes the reinvestigation of possibly prematurely dismissed solutions, such as the luminescent solar concentrator (LSC). In this research an inorganic thin film LSC is characterized experimentally in detail in terms of all the separate light transport steps that result in the concentration of sunlight. A 3 um thin film of Eu3+ doped Lu2O3 was chosen for of its large Stokes' shift, which excludes all transport losses due to self-absorption and allowed to study losses caused by scattering at interfaces of the LSC. A model is presented which can be used to calculate the LSC light transport efficiency as a function of LSC surface area. This model needs the easily measured linear attenuation as input, which characterizes the transport efficiency. A second more elaborate analytic model is developed to separate the losses in the film and in the substrate. This model also considers the non-uniformity of the attenuation length. The quantum efficiencies of our LSCs following from the measurements are in the order of 15%, which is less than ideal, mainly due to a poor 34 - 44% luminescence quantum efficiency and a 53 - 65% waveguide efficiency. Measurements have shown some discrepancies. Both the absorption spectrum and the time resolved luminescence spectrum did not show the expected behaviour upon change of europium concentration. The directional output measurements did not show the expected relative intensities of the film and the substrate. Qualitatively, however, the model and measurements are similar. Complimentary measurements could reveal the causes of these quantitative differences. A possible reason could be that some waveguide modes are not supported by the film. The modelling reveals that, in order to have building integrated LSCs simultaneously acting as windows, the linear attenuation length in the LSC should be longer than one meter. This is in contrast to the values in the order of tens of millimetres, which have been measured for the LSCs in this work. If such attenuation lengths are realised, power efficiency calculations reveal power efficiencies of up to 16% for the appropriate combination of luminescent material and solar cell.