Analytical quality by design (AQbD) principles are well introduced in pharmaceutical industry and have big influence on analytical methods because of the high importance of quality control in this industry. These principles are adaptable in environmental analysis also and can be integrated into environmental quality management system. In the present work an attempt is made to use the principles of Quality-by-Design in the process of analysis of oil compounds in soil. The analytical target in this case is oil compounds that are spilled into soil, namely 16 parent polycyclic aromatic hydrocarbons (PAHs) and their alkylated analogs, 7 polychlorinated biphenyls (PCBs), and three aliphatic fractions from decane to pentatriacontane. The method used was gas-chromatography mass spectrometry (GC-MS) where the design of the experiment (DoE) was applied, in order to choose the optimal chromatographic conditions. The chemometric methods (Pearson product-moment correlation, Shapiro-Wilk test, and the principal component analysis) were applied for a comparison of methods used for analysing PAHs in soil, and the identification of the soil samples that were polluted with diesel oil and motor oil. The developed method was evaluated with respect to linearity, the limit of detection (LoD), and the limit of quantification (LoQ), within laboratory precision and the measurement uncertainty. Cross-validation was undertaken to compare the common laboratory method of PAH analysis and a novel analysis using the t-test. The Z scores were calculated when using the results of the proficiency test for 16 PAHs and 7 PCBs. A combined modified procedure based on ISO 18287:2006 for the soil quality, together with the SPIMFAB method as an instruction for analyses of the aromatics fraction > C16-C35, was proposed for the two matrices (soil and sediment). The results show that the recovery rates for the spiked samples were nearly 100% for both the soil and the sediment.