The current investigation, conducted in the general framework of the MUSE program ("MUltiplication avec une Source Externe"), considers the representativity of a specific configuration of its fourth phase (M4SC2), which is driven by an external D(d,n)He3 or T(d,n)He4 neutron source, with respect to current concepts of eXperimental Accelerator Driven Systems (XADSs) with gas (He), Na and Pb/Bi coolants. The study has been carried out from the nuclear data viewpoint, with the external source being accounted for in an appropriate manner. In this context, data sensitivity/uncertainty analyses based on first-order perturbation theory calculations have been performed using the deterministic code ERANOS (Version 2.0) in conjunction with its adjusted nuclear data library ERALIB-1. It is found that the M4SC2 configuration, independent of the external source, is quite representative of the different XADSs for actinide capture reactions at the centre of the fuel zone, relative to 239Pu fission at the same location. For the case of a threshold fission reaction, such as that in 238U, the sensitivity to the external source is significantly higher. With respect to the corresponding spectral index, M4SC2 with the D(d,n)He3 source remains quite representative of the He- and Na-cooled XADSs. For the system with Pb/Bi coolant, on the other hand, effects of uncertainties associated with the data for these two nuclides and their low content in the MUSE configuration result in significantly lower associated representativity factors. A better overall representativity of the Pb/Bi-cooled XADS is expected to be achieved by the new MUSE_Na/Pb configuration. [All rights reserved Elsevier]