Approximate half a million people in Germany are treated with anticoagulants. It is expected that the market for antithrombotics will double until 2010. The currently used anticoagulants have disadvantages such as a lack of oral availability or narrow therapeutic windows. Natural products have been a rich source of lead structures. Among the group of anticoagulants there are many agents which derive from natural products. Within the different classes of natural products, plants offer the easiest way to isolate the active substance by upscaling. Therefore plants were the starting point for our search for thrombin inhibitors. The aim of this work was to isolate thrombin inhibitors from plants. In a first step methanol and dichlormethane extracts of 78 different drugs were prepared. They were tested for thrombin inhibition using an amidolytic assay. In order to avoid an inhibition due to the presence of certain substance classes we tested on tannins, macromolecules and promiscuous inhibitors. At the end of all tests, Adonis herb was chosen for futher examination. By testing the inhibition of thrombin before and after removing tannins, we can determine if the inhibition was only caused by tannins. Due to the great number of samples, the use of a precipitation method for the removal of tannins is to be prefered. Common as precipitation reagents are caffeine, polyvinylpyrrolidone, gelatine and lead acetate. The pH, different solvents and the concentration of the precipitation reagent were compared for their influence on these four tannin precipitation reagents. Futhermore the influence of the precipitation reagents on the activity of thrombin and the precipitation of other phenols was examined. Polyvinylpyrrolidone showed the best properties to remove tannins under the test conditions used. Lead acetate proved to be inappropriate. Tannins are usually said to have a high reactivity linked with an unspecific inhibition. With the aim of proving this thesis, it was tested if condensed tannins interact with different proteins selectively. Thereto the condensed tannins from Cinnamon bark were isolated. Then the condensed tannins were precipitated with different proteins. After the precipitation the condensed tannins were degradaded and the degradation pattern were compared. It was shown that condensed tannins interact with different proteins selectively. Moreover, the compound in Adonis herb which causes thrombin inhibition was found and it was shown that the fatty acids linoleic and linolenic acid inhibit thrombin. Additionally palmitic acid was found in Adonis herb. Two other drugs, Erucae semen and Sambuci flos, which also inhibit thrombin, were examined as well. It was shown that again fatty acids inhibit thrombin. Linoleic, linolenic, oleic and eicosenic acid show inhibition. A mixture of the fatty acids did not show a synergistic inhibition. A futher analysis of the inhibition of thrombin by fatty acids showed that vaccenic, eicosatrienic, ricinoleic and arachidonic acid inhibit thrombin as well. However, the inhibition is time-dependent and dependent from the incubation time. Dependency on the incubation time occurred only when the amidolytic assay was started by the addition of the substrate. Because of futher experiments a blockade of the active site by fatty acids seems unlikely. A connection between the inhibition of thrombin and the physicochemical parameters of the fatty acids could not be established. The common procedure for the isolation of inhibitors from plants is a bioassay-guided isolation. Since several years the method "Ligandenfischen" is used in our working group to accelerate the bioassay-guided isolation. Similar methods were published as well. All methods share the separation of an enzyme-ligand-complex from the matrix. By subsequent analysis of this complex, futher information like the retention time and the uv spectra can be obtained. These pieces of information could accelerate the process of the isolation of a compound. As another part of this work it was shown that the interaction between thrombin and tannins can be used to separate an inhibitor as enzyme-inhibitor-complex from a matrix. The method was simple, cheap, fast and practicable by means of analytical hplc.