Food matrices consist of many components with different physical and chemical properties that may influence instrumental robustness. The soybean contains fatty coextractives which may have a deleterious effect on the gas chromatography (GC) system. In this study, the efficiencies of PSA, C 18 OH, C 18 , silica, aluminum oxide, and Florisil, as dSPE clean-up sorbents, were evaluated by the high-performance liquid chromatography (HPLC) diode-array detector and evaporative light-scattering detector analysis. The dithiocarbamates in soybean samples are determined as CS 2 using acidic hydrolysis and isooctane partitioning, followed by GC-PFPD and GC-ITD-MS analyses. The linearity of the analytical curves, the instrument limit of detection matrix effects, the trueness and precision, and the method limit of quantification (LOQ) were assessed in the validation study. Milled soybean was spiked with thiram solution at three concentration levels (corresponding to 0.05, 0.1, and 0.5 mg CS 2 kg −1 ) for recovery determination. Silica appeared to be an effective and cheap sorbent to remove coextracted matrix components without causing any CS 2 losses. Recoveries were in the range of 68−91%, with relative standard deviations ≤ 8.7%. The method LOQ was 0.05 mg CS 2 kg −1 , and both GC-ITD-MS and GC-PFPD systems appeared to be appropriate and complementary to determine dithiocarbamate residues in soybean extracts.