The development of varieties resistant to green peach aphid (GPA) faces difficulties when doing the accurate identification of resistant genotypes. The use of physical and chemical leaf surface characteristics as selective criteria in practical breeding is very contradictory. However, the most important method of the identification of resistant genotypes still remains the evaluation of plant infestation in field tests. The aim of the presented work was to monitor the heritability of resistance to GPA and to test the appropriate selection procedure in the development of resistant lines. Field tests of varieties, F1 hybrids, F2 population, backcross and pedigree lines were conducted in the neighbourhood of anticipated presence of aphids. Estimates of plants were carried during the first pressure of aphids, the stabilisation of adult aphids population and at the end of vegetation. The evaluation of selected F1 hybrids gained from the crossing of resistant varieties TI 1112, TI 1028, TI 1025, TI 1132, Vy 32 with susceptible varieties NC 2326, V 130, V 129 showed that there exist a recessive character of resistance to GPA. F1 hybrids originated from the crossing of TI 1112, TI 1028, TI 1132, TI 1025 with NC 744, F 225, Vy 32 genotypes maintained their genetic resistance to GPA except of TI 1112 which in combination with resistant genotypes made susceptible F1 hybrids. It might be concluded that TI 1112 variety, often used as a model genotype, has different genetic background. Eventually, the susceptibility of the above F1 hybrids is connected with gene that controls secretion activity. The analysis of progeny of resistant and susceptible genotypes found in F2 generation revealed fixation of the resistance and the accomplishment of homozygotic state in the first year. Testing the plants under high natural infestation pressure with the use of alternative estimation method appears as the appropriate selection procedure in the development of resistant tobacco cultivars.