In this paper we describe a new approach to sampling-based motion planning with Probabilistic Roadmap Planner (PRM) methods. Our aim is to compute good quality roadmaps which encode the multiple connectedness of the configuration space inside small but yet representative graphs which capture the different varieties of free paths well. The proposed Path Deformation Roadmaps (PDRs) rely on a notion of path deformability indicating whether or not a given path can be continuously deformed into another existing path. By considering a simpler form of deformation than that allowed between homotopic paths, we propose a method that extends the Visibility-PRM technique to construct compact roadmaps which encode richer and more suitable information than representative paths of the homotopy classes. PDRs contain additional useful cycles between paths in the same homotopy class that can be hardly deformed into each other. Experimental results show that the technique enables small roadmaps to reliably capture the multiple connectedness of complex spaces in various problems involving free-flying and articulated robots in both two- and three-dimensional environments.