Untargeted liquid chromatography-high resolution mass spectrometry (LC-HRMS) techniques have become indispensable tools for systematic toxicological analysis. They allow the research of an almost unlimited number of drugs within a single analytical cycle, but shared mass spectra libraries are still missing to identify newly marketed compounds, along with defined analytical procedures. This article describes the optimization, validation, and application of an untargeted screening method devoted to hair analysis using data-dependent analysis (DDA) and a shared HRMS database. This method used an ultra-high performance liquid chromatography coupled to a benchtop Orbitrap. Raw MS data were processed with Compound Discoverer software coupled to the mzCloud™ library. Optimizations were performed on blank hair spiked with 19 analytes having different physical and chemical properties. To validate the effectiveness of a shared spectra database, 20 compounds spectra were added and then retrospectively screened.Sensitivity and reliability were evaluated on 317 compounds of interest in toxicology. The method was then applied to 11 hair samples. The matrix effect range by ion suppression/enhancement was 40%-110%. The method allows the detection of 284 among the 317 screened compounds, including 72 new psychoactive substances (NPS).Lower limit of identification (LLOI) and lower limit of detection (LLOD) were 1 to 1000 pg/mg and 1 to 500 pg/mg, respectively. The method was successfully applied to 11 clinical cases and 144 compounds were identified including 24 NPS including AKB48-5F for the first time in hair. We developed and validated an LC-HRMS untargeted screening of 284 compounds and successfully applied it to 11 real hair samples.