In part 1 of this paper, the triaxiality function of several adhesive joints was investigated using the finite element method. The main aim of this investigation was to decide on the most suitable joint type for fatigue experimental tests, which can be used to control triaxiality in adhesive layers. These fatigue experimental tests are presented in this paper (part 2). In part 1 of this paper, three types of adhesive joints were considered. These were: butt joint, cleavage joint and scarf joint. The geometry of the butt joint cannot be manipulated in order to control triaxiality, and therefore triaxiality is controlled through the tension/torsion loading ratios. It was found that the type of load had an effect on the map of triaxiality in the adhesive layer. For the cleavage joint, it was found that triaxiality can be manipulated by changing the adhesive bondline angle. However, the relationship between triaxiality and bondline angle was found to be complex. For the scarf joint, the triaxiality can also be manipulated in a similar way to that used for the cleavage joint by changing the bondline angle. The scarf joints has many advantages over cleavage joints including: (a) simpler relationship between triaxiality and bondline angle and (b) easier manufacturing process related to the geometry of the joint. It was shown that the scarf joint is the best practical choice to control triaxiality in fatigue experimental tests. (C) Koninklijke Brill NV, Leiden,