Grapevine is grown as a grafted plant, mainly using phylloxera-resistant rootstocks obtained when this aphid destroyed European vineyards, and the use of a reduced number of rootstocks in each production area is common. This indicates that the genetic variability that is being used could be insufficient to tackle new stress constraints. Changes that will be produced as a consequence of climate change are promoting the development of new rootstocks and the study, in a deeper manner, of those already in use, mainly in relation to drought stress. In this work, we have studied 40 rootstock accessions, including clones of common rootstocks, others developed later, some recovered from old abandoned fields and other, resprouted rootstocks. From these accessions, 19 unique SSR profiles were obtained and chlorotypes were assigned, as no information was available for them in the VIVC database, thus generating new knowledge. Genetic variability was analysed in the 110 Ritcher, 140 Ruggieri and 1103 Paulsen rootstocks (derived from Vitis berlandieri and Vitis rupestris), commonly used in the countries of greater wine production (Spain, France and Italy), and in the 19 rootstocks with unique profiles. As expected, higher variability was found in the latter. Fortunately, variability was also found in the small sample of which reflects there is variability among the three more-commonly-used rootstocks despite they are half and/or full sibs. Considering all the germplasm analysed, the relationships found agree with a recent report stating that some genotypes had been erroneously assigned, previously, and show that another genotype may not be correct. Variability was also found in clones of several rootstocks, with considerable variability in some of them, including two rootstocks rescued from old abandoned vineyards. This result suggests the possibility of evaluating these materials for other traits. Finally, evaluation of osmotic-stress tolerance was carried out in in vitro culture, using media containing PEG. Micropropagated plants of one rootstock classified as drought-resistant, another reported as sensitive and two others whose classification in the field is variable were used. The results indicate that this methodology can be useful in breeding programmes, to screen the variability in osmotic-stress tolerance among clones and to study root architecture and plasticity.