Target screening analysis of environmental samples is limited to a list of priority pollutants omitting other potentially hazardous substances or transformation products. In nontarget screening, no prior selection of micropollutants is made, which gives all compounds initially the same chance of getting detected. Which ones eventually are, depends on the choice of analytical instrumentation and the way the acquired data is processed and analysed. In this project, a set of workflows for nontarget screening of solid environmental samples was developed. The aim was to use a sample preparation strategy with minimal discrimination of compounds; to develop complementary analytical methods that can separate and identify compounds with a wide range of polarities and volatilities; to apply pixel-based multivariate statistical tools to characterise urban freshwater sediments based on their chemical fingerprint; to test several workflows for compound and source identification. It is crucial to understand potential sampling errors to conduct soil and sediment sampling in a representative manner. Currently, if soil needs to be removed due to constructions, Danish law requires that one soil sample is collected every 30 tonnes and classified based on the concentration of specified target pollutant. The evaluation of a GANDALF soil sampling campaign in 2018 (sampling grid size of 7×7 m) showed high sampling uncertainties for all targeted compounds, e.g., 222.1 % for oil hydrocarbons (C15-C20). Predictions showed that changing the sampling grid size (e.g., between 1 and 10 m) would only marginally affect the sampling uncertainty (ranging between 215.1 and 225.3 %). The more significant effect was based on increasing the number of increments, i.e., single samples that are combined to forma composite sample.To be able to extract a broad range of micropollutants with a varying polarity, a sequential pressurised liquid extraction was developed for sediment and soil samples. The first step was a extraction at 50 °C aiming at polar compounds with methanol:water [1:1]; the second step, at 100 °C aimed at extracting semi- and nonpolar compounds with dichloromethane:acetone [3:1]. The extracts from the second extraction step were analysed with gas chromatographymass spectrometry (GC-MS). Recoveries of target analytes (0.3 < log P < 6.9) ranged between 20.9 and 116.6 % in the second extraction step. The evaluation of the first extraction step and nonvolatile analytes in the second step extract is ongoing. Nontarget screening with chromatographic and MS tools should provide a high peak capacity, mass resolution and accuracy so that unequivocal identification of many compounds is possible. Comprehensive two-dimensional GC (GC×GC) and high-resolution (HR) MS were applied to obtain chemical fingerprints from sediment samples (Manuscript II). For the first time, a two-dimensional pixel-based principal component analysis was applied to urban freshwater sediments to identify chemical pollution sources. The work also highlights that careful considerations of the data pre-processing steps (i.e., retention time alignment, noise reduction and normalisation) are necessary to reach meaningful conclusions reflecting the chemical differences between the samples without unwanted and interfering variation. Complementary nontarget screening methods for sediments and unconventional oils weredeveloped with supercritical fluid chromatography (SFC)-HRMS to extend the range to nonvolatile and more polar compounds compared to GC (Paper I, II, and Manuscript I). In Paper I, the retention behaviour of polycyclic aromatic compounds and their oxygenated derivatives was examined. A column screening of three stationary phases and solvent modifiers showed that both the DIOL and 2-PIC column retain polar compounds more than the BEH column. The highest peak capacity was achieved with methanol and the 2-PIC column, and peak shapes were improved with the addition of 0.1% formic acid to the modifier. In Paper II, different processing conditions for the production of pyrolysis oil feeds were linked to the type of oxygenated compounds. Pixel-based PCA helped to prioritise peaks for identification (e.g., higher fatty acid contents in the light feeds compared to the heavier feeds). The evaluation of the sediment dataset that was analysed with SFC is still pending, but a few preliminary results are presented in the thesis. The importance of carefully designed experiments was shown based on the use of Doptimal algorithms to optimise the SFC chromatography and MS sensitivity with electrospray ionisation (ESI), based on 60 standard analytes. Key results were that the use of a shallow gradient slope (1 %B min-1 ) is essential to obtain a high peak capacity (> 300), whereas the concentration of modifier additive (here ammonium formate) and water, capillary voltage and make-up solvent flow rate have a significant effect on the ionisation efficiency (Manus-cript I). Numerous tools are available for compound identification workflows, e.g., databases and suspect screening lists, diverse in silico fragmenters, tools for data reduction and prioritisation, and communication of identification confidence. The pixel-based PCA models were used to prioritise the compound identification to the components in a sample with the highest variability (Paper II and Manuscript II). Based on tentatively identified compounds in the GC×GC-HRMS data (Manuscript II), it was possible to identify chemical pollution sources in the urban freshwater sediments from a weathered petroleum-based oil (e.g., based on alkylated benzenes, cyclic aromatic hydrocarbons and naphthalenes), a natural background (e.g., plant-based oils such as mono-, di- and sesquiterpenes) and road runoff (e.g., benzothiazoles). The use of retention indices increased the level of confidence for many tentatively identified compounds. Further, several homologous series were identified in a homolog screening of SFC-ESI-MS sediment data, including anionic surfactants (linear alkylbenzene sulphonates) and previously not reported dialkyl- and dialkylmethyl-amines which are industrial additives for many applications and were GHS classified as very toxic to aquatic life. The analysis and evaluation of the SFC-ESI-MS sediment data as well as the compound identification based on the in silico fragmenter MetFrag are still pending.