Within this thesis, a LC-MS based method was developed, which enables the identification of activity relevant metabolites in complex mixtures such as natural product extracts prior to isolation. Therefore, a direct activity-correlation-analysis (AcorA) has been developed and adapted to metabolite profiles and the corresponding (bio-)activity profiles. As basic principle, a proportional relationship between the peak intensity and biological activity is assumed. As biological activity, the antibacterial inhibition of the gram-positive bacterium Bacillus subtilis was chosen. Secondary metabolites from the fungal genus Hygrophorus (basidiomycetes) are promising candidates for antibacterial compounds. Thus, several species of this genus were analysed using AcorA, a LC-MS3 method, and a fluorescence-based antibacterial bioassay to assess the biological activity of the extracts. AcorA enabled an ab initio identification of the responsible bioactive compounds within the crude fungal extracts without any purification. For verification, the metabolites were isolated by a m/z-guided strategy and structurally characterized by different mass spectrometry methods. The bioactive compounds were identified as dihydroxylated, unsaturated fatty acids and an oxidized cyclopentanone derivative, which all show antibacterial activity against B. subtilis.