Compared with the primary metallurgy the production of aluminum from secondary raw materials has many advantages, like lower energy consumption, a smaller amount of resulting residues, avoidance of landfill capacity etc. For an Al-recycling company the recycling process starts with the delivery of scrap and ends with the sale of the final good. A downcycling of the material during the process should be avoided. Since the possibility for refining aluminum is limited, an accurate characterization of the starting materials is necessary to prevent the input of harmful elements from the scrap into the product. Another important point regarding the profitability are the losses of metal during the process. The tasks of this work are given by the mentioned facts. An important point is the sampling of scrap and the detection of the content of volatile components (water and hydrocarbons). Possible ways for measuring these components and their potential effect on the product (entry of unwanted substances) and the process (formation of volatile hydrocarbon compounds, reducing the metal yield) are discussed in the theoretical part of this work. In addition, a new concept for an industrial measurement device to detect the content of organics in Al-scrap is given. Extensive melting experiments were made to investigate the effect of a scrap pretreatment (mechanical and thermal) and the existing melting technology (with or without melting salt) on the metal recovery. In terms of melting loss there was a detailed analysis of the factors influencing the yield at rotary- and tilting rotary drum furnaces. Although the effect of some of the relevant factors which have been examined (e.g. the addition of fluoride, the salt composition and quantity of salt) is already known, further investigation of this topic was necessary, because the available information are based on laboratory tests. Therefor the parameters at this experiments may not always the reflect reality. The aim of this thesis was to investigate the effect of the influencing parameters under production-related conditions and the alienability of the results to the industrial production process.