The aim of this diploma thesis was to find an alcohol dehydrogenase ADH, which accepts primary alcohols as substrates. As a standard substrate benzyl alcohol was used, which is oxidised to form benzaldehyde. At the same time the cosubstrate acetaldehyde is reduced to form ethanol, whereby the nicotinamiddinucleotide NAD+is recycled. After a long search for suitable microorganisms, Janibacter terrae DSM 13953, was identified as a promising candidate. For purification PIPES was used as a buffer, for pre-cleaning, the cell wall of J. terrae had to be pretreated with lysozyme, so that more protein from the cells could be released during subsequent ultrasonic lysis. This was followed by a temperature and ammonium sulfate precipitation to remove cell debris and other protein from the supernatant. After the completion of the pre-cleaning the actual purification and concentration of the protein could be started. For this task, hydrophobic interaction chromatography HIC, ion exchange chromatography IEX, affinity chromatography with BlueSepharose 6Fast flow and size exclusion chromatography SEC were used. Purification was checked by SDS electrophoresis. When only one band was left, it was excised and sequenced to perform a homology search in the genome of J. terrae and other known proteins. Unfortunately no hits were found which leads to the conclusion that the given genome was inappropriate, because the search with blast provided many ADHs. Of the two best matches, an NDMA oxidoreductase from Amycolatopsis methanolica and an iron-containing ADH from Nakamurella multipartita DSM 44233, vectors containing a Strep-Tag were prepared and the enzymes were tested for activity. Unfortunately no activity was detected for the two proteins, neither for the standard substrate benzyl alcohol nor for the natural substrates of the two ADHs, methanol and ethanol, which probably points to a denatured protein (caused by the Strep-tag or another factor).