The successive stress-relaxation (SSR) experiment is a well- known technique that is mainly used to study different aspects of the plastic deformation phenomena. The technique has been described extensively elsewhere [L.P. Kubin, Philos. Mag. 30 (1974) 705; P. Sp¨atig, J. Bonneville, J.- L. Martin, Mater. Sci. Eng. A167 (1993) 73; J. Bonneville, P. Sp¨atig, A. Orlov´a, J.-L. Martin, Proceedings of the 10th International Conference on the Strength of Metals and Alloys, Japan, 1994, pp. 199–202; J.-L. Martin, T. Kruml, J. Alloys Comd. 378 (2004) 2; J.-L-Martin, B. Lo Piccolo, T. Kruml, J. Bonneville, Mater. Sci. Eng. A322 (2002) 118]. In this paper, we briefly review the assumptions in which the analysis of test results is based, within the framework of thermally activated processes. In particular, we review the hypotheses usually invoked regarding the evolution of the mobile dislocation density (ρm) during the overall duration of the SSR experiments. In particular, the SSR data analysis assumes that there is no increase of ρm during the reloading between two successive relaxation transients (quasi-elastic loading). We propose a method to check the self-onsistency of this hypothese. © 2007 Elsevier B.V. All rights reserved.