A fully automated screening using liquid chromatography-mass spectrometric method applying data-dependent acquisition was developed to identify toxicologically relevant substances in serum and urine. A library including more than 405 spectra of about 365 compounds (main drugs and important metabolites) was established. An easy to use program was created to automate and accelerate library search. Drugs were identified based on their relative retention times, molecular ions and fragment ions. Limits of detection were tested with 100 of the 365 compounds the majority of these were lower than 100?g/l (67%). The developed LC-MS-MS system seems to be a valuable alternative to other general unknown screening methods allowing fast and specific identification of drugs in serum and urine samples.