The aim of performed investigation was to evaluate the influence of changes in herbicide application system on herbicide residues in soil and sugar beet roots. Chemical weed control in sugar beet was carried out by herbicides that included substances such as phenmedipham, desmedipham, ethofumesate, metamitron, triflusulfuron and surfactant adjuvant applied in three different systems: two times application at bare soil (preemergence) and postemergence application (weeds in phase of 2–4 leaves) – system “A”, 3 times split, postemergence application (full dose of herbicide mixture) – system “B” and 4 times application at 7 to 10 day intervals starting at the beginning of weed emergence – system “C”. Samples of soil and roots of sugar beet were taken at the day of lifting. Herbicide residues were analysed using HPLC with UV-detection. At lifting time, in soil samples, where herbicides were applied in system “A”, the residues of metamitron amounted from 0.0097 to 0.0132 mg/kg. Sum of all detected residues of applied substances amounted 0.0341–0.0458 mg/kg. In sugar beet root samples, the residues amounted to respectively, 0.0049–0.0064 and 0.0136–0.0247 mg/kg. The application of herbicides in “B” and “C” systems caused a significant decrease of residues by about 50% (system “B”) and 65% (system “C”) on average, in comparison with results obtained for herbicide application in “A” system. Residues of active substances determined in roots of sugar beet did not exceed acceptable limits (MRLs).