A successful exploitation of in vitro tools for breeding and for the understanding of gene functioning requires the regeneration of true-to-type plants and experiments were therefore performed with several genotypes of Medicago truncatula (J5, TRV25, TR122) to characterize mother plants and regenerating tissues. Each sample was assessed by flow cytometry and, whatever the genotype and regeneration pathway, a divergent phenotype was systematically linked to an abnormal nuclear DNA content. All samples assessed were classed according to their flow cytometry profiles into normal (true-to-type) material, aneuploids, endoreduplicated tissues, tetraploids, mixoploids and senescent tissues. Deviating calluses failed to regenerate or gave rise to infertile, non-viable plants. In turn, all tissues with non true-to-type flow cytometry profiles were examined in terms of isoenzyme banding patterns compared to the mother plants. Esterases, Peroxidases and Leucine aminopeptidase appeared to be the best isozyme systems to show differences between the original genotypes but also between diverging materials and the mother plants. Interestingly, such differences were more often qualitative (presence or absence of bands) than quantitative (i.e. differences in colour intensity of bands) thereby making easier an accurate distinction between genotypes. Peroxidases were prone to variation with culture medium and tissue age. The results stressed the importance of using more than one approach when undertaking the characterisation of materials as, for some of the genotypes analysed, differences compared to the respective mother plants could be shown with flow cytometry that were not reflected in a different banding pattern with isoenzymes.