The presence of joints in rock masses influences the structural integrity of geotechnical structures. A critical failure mode is shearing, thus making the shearing process of importance to understand. Historically, studies have been mainly executed on the basis of laboratory experiments, since full-scale in situ tests are seldom performed due to technical and economic considerations. Since each rock joint is unique by nature, the utilization of replicas is applied to carry out controlled experimental parameter studies. However, the manufacturing process of replicas introduces many sources of uncertainty. Therefore, in this work the influence of geometrical variations in replicas on the shear strength characteristics is evaluated, mutually as well as in relation to the mother rock specimen of the replicas. The joint surfaces were 3D scanned and the contact area of the joint was measured using pressure sensitive film before direct shear tests. Deviations in morphology were evaluated by surface comparisons between the joint surfaces of the mother rock and replicas. The initial matching of the joints was evaluated by calibrating the scanning data with respect to the contact area measurements. It could be visualized that geometrical deviations were caused by rock fragments coming off during mould production, positioning of the moulds and pores resulting from replica casting. These factors were found to influence the shear strength characteristics of the replicas. The influence of the deviations originating from morphology on the joint matching is demonstrated. In summary, it is shown that replicas with similar shear strength characteristics as rock can be manufactured, but even small deviations affect the characteristics, in particular the peak strength. Therefore, parameters relevant for geometrical quality assurance should be identified along with required value ranges. Selected introductory results on quantified parameters for geometrical quality assurance are presented, serving as a basis for continued work.