Quantifying microplastic pollution, i.e. particles smaller than 5 mm, in complex environmental samples has become a fundamental issue for the assessment of the environmental and health consequences of this pollutant. Pyrolysis-GC-MS emerges as a promising alternative, and for robust quantification, it is essential to improve our understanding of the aspects interfering in the pyrolysis of microplastics. In this work we investigate the effects of the polymer molecular weight and soil matrices on the pyrolysis-GC-MS analysis of polystyrene (PS) microplastics. We also evaluated the effectiveness of the addition of Poly(4-fluorostyrene) (PSF) as an internal standard to minimize matrix effects. PS microplastics (of low - 35.10³ and high - 400.10³ molecular weight, LMW and HMW, respectively) were individually spiked in different matrices and calibration curves were prepared in the range 0.1 to 2.5 mg/g. Samples were directly injected into the pyrolysis-GC-MS system and the three main markers of PS pyrolysis (styrene, styrene dimer and styrene trimer) were considered. PSLMW showed a reduced formation of all markers compared to PSHMW, and in both cases the styrene was proportionally more formed than the other markers. Due to differences in the slopes of calibration curves, quantification errors up to 5-fold might occur depending on the marker chosen and the polymer molecular weight used to build the curves. A strong matrix effect was noticed, sharpening the answer of PS markers in a 2.6 to 7.5-fold range. This effect was mostly attenuated through the normalization by PSF markers. In the absence of an internal standard, external calibration curves should be avoided, and the standard addition method might be applied. In addition, the development of reference materials for the study of microplastic pollution should encompass the complexity of the target analyte, including the diversity of molecular structures, such as oxidation degrees and molecular weights.