For the purpose of integrated plant protection, it is of high relevance to include recent and future innovations into plant protection management. A site-specific management of plant diseases will lead to a reduced input of agrochemicals and thus to ecological and economic benefits. Previous research showed that hyperspectral sensors can be used to detect and quantify crop stress that is caused by plant diseases. The majority of these studies were conducted under controlled conditions because hyperspectral sensors demand high standards to the environmental conditions and the measuring system. These high demands make a field approach challenging as there are many parameters to consider to generate high quality, reproducible and comparable field data. This work presents the main factors that affect the quality of hyperspectral field data and shows basic approaches to develop protocols for hyperspectral field measurements. The requirement for a reliable disease detection in the field is to produce comparable field data on different measurement days with the same setup. Therefore, different approaches were tested and used. The gain of knowledge out of hyperspectral measurements on different scales and under differing environmental conditions has to be integrated through scale transfers and recent analysis methods, such as feature selections, supervised classifications and regression analysis. Main goal is the definition of the characteristics for one sensor that fulfills the practice demands for plant disease detection under uncontrolled field conditions.