Abstract: The study aimed to evaluate distances required for oilseed rape to prevent outcrossesbetween different cultivars and also between neighboring weeds of Brassicaceae. Evaluation ofpossible outcrosses of traditional cultivars growing in close proximity is a good model to estimatethe probability of gene flow from GM plants. Modelling was made by field experiments,combining them with molecular analysis and pollen study. Five cultivars (‘Maskot‘, ‘Sw Savan‘,‘Heros‘, ‘Ural‘, ‘Landmark‘) of spring oilseed rape (Brassica napus L.) were grown inexperimental plots in Lithuania. The number of weeds was registered in plots established parallelto the different cultivars. Plastid SSR primers MF1, MF2, MF3, MF4 were used for the molecularexperiments. Amplification was successful with all used primers; two SSR loci used in our studywere polymorphic after amplification with primers MF2 and MF7. Results of oilseed rape geneticdiversity showed a significant genetic variation among individuals and different cultivars.Estimations were done concerning dispersion of pollens of oilseed rape by bees. The amount (asmass proportion, estimated in %) of oilseed rape pollens was 66 and 92 % in two bee hives placednear B. napus cultivars at the beginning of the study period (July 12) and decreased to 24 and 49% until July 28 in the two hives, respectively. Examined cultivars significantly differed in respectto the concentration of N, Ca, Mg, and S on the pollen surface. Our preliminary data suggest thepossibility to use pollen surface element analyses for the identification of specific cultivars.