Bone morphogenetic protein-2 (BMP-2) is considered a promising adjuvant for the treatment of bone regeneration. However, BMP-2 delivery in a conventional collagen scaffold needs a high dose to achieve an effective outcome. Moreover, such dosage may lead to serious side effects. The aim of the following thesis was to find clinically acceptable strategies reducing the required dose of BMP-2 by improving the delivery and optimizing the preclinical testing of the new approaches. In all the studies hyaluronic acid (HA) hydrogels was used as a carrier for BMP-2. The HA hydrogel/BMP-2 construct was modified with bioactive matrix components in order to obtain an effective release of BMP-2 and an enhanced bone formation. The most promising were two strategies. In the first one, BMP-2, precomplexed with the glycosaminoglycans dermatan sulfate or heparin prior to loading it into HA hydrogel, protected and prolonged the delivery of the protein, resulting in twofold larger bone formation in comparison to non-complexed BMP-2. In the second strategy, the fibronectin fragment integrin-binding domain (FN) was covalently incorporated into HA hydrogel. The FN remarkably improved the capacity of the material to support the cells attachment and spreading, providing the formation of twice as much bone in comparison to non-functionalized HA hydrogel/BMP-2. Furthermore, the importance of a proper design of the preclinical study for BMP-2 delivery systems was highlighted. Firstly, proper physicochemical handling of BMP-2 showed the improvement in further in vivo activity.  The use of glass storage vials and an acidic formulation buffer was superior to plastic surfaces and physiological pH. Secondly, while regenerative medicine strategy testing required the use of animal models that matched the research questions related to clinical translation, two new animal models were developed. The subperiosteal mandibular and calvarial models in rats were found to be minimally invasive, convenient and rapid solution for the evaluation of a broad range of approaches including bone augmentation, replacement and regeneration. Both models are primarily relevant for the initial testing of the injectable bone engineering constructs.  Those clinically translatable approaches presented here could prove to be a powerful platform for a wider use of BMP-2 in orthopedic, plastic surgery and regenerative medicine research.