Load-bearing tissues such as articular cartilage or meniscus are not infallible. They might fail due to trauma, over-solicitations, or diseases, which could subsequently alter the mobility of the patient. Repairing these damaged tissues in a minimally invasive way is nowadays widely studied. Hydrogels are promising for repairing soft tissues. However, trying to reproduce the highly hierarchical microstructures of load-bearing tissues is extremely complicated. In fact, conceiving hydrogels with optimal properties for specific applications is challenging as most of the hydrogel's properties are interrelated, such as toughness, stiffness, swelling, or deformability. The improvement of one property usually comes at the cost of another. This thesis aims to develop different hydrogel microstructures and evaluate their potential for load-bearing implants. In particular, the study focuses on (i) designing tough and fatigue resistant hydrogels, (ii) establishing a methodology to control the stiffness and the swelling of hydrogels independently, and (iii) considering processing ease by tailoring each hydrogel precursor viscosity. An extensive parametric study on hydrogels' structure and composition allowed establishing complete property charts for mechanical, swelling, and rheological properties. Seven different hydrogel structures based on poly(ethylene glycol) dimethacrylate were developed: neat, double network, composite, double network composite, granular, hybrid granular, and silk granular hydrogels. The precursors of neat hydrogels had the lowest viscosity. Microgels of similar composition were added to the precursor to adapt and increase its viscosity for unconfined applications. In parallel, adding nano-fibrillated cellulose fibers was effective for increasing toughness. This method was more efficient than creating a double network with alginate. Moreover, fatigue tests revealed that hydrogel composites successfully survive 10 million loading cycles at 20\% applied strain. However, it became softer after the first loading cycles and behaved similarly to the Mullins effect. Subsequently, we assessed how cyclic loading affects fracture behavior, distribution of strain fields, and microstructure. The study demonstrated that cyclic loading on hydrogel composites re-arrange the fiber network without seriously deteriorating the mechanical properties. Additionally, we demonstrated that combining composite and microgel approaches, i.e. hybrid granular hydrogels, effectively tailored hydrogels' swelling and viscosity without significantly affecting stiffness, toughness, and deformability. Finally, the microgels were swollen in silk fibroin solution for forming self-reinforced silk granular hydrogels. Those hydrogels exhibited considerably higher stiffness and lower swelling but also reduced elongation performances. We demonstrated that hydrogels' mechanical and physical properties could be effectively controlled with the microstructures and compositions of well-known biomaterials. Subsequently, a hydrogel composite and a silk granular hydrogel, based this time on gelatin hydrogels, were synthesized, validating the potential of studied structures with other hydrogels and microgels. Finally, the analyzed material systems can be combined through a sequential layering, forming hybrid hydrogels, to obtain local or gradient properties addressing specific biomedical applications, soft robotics, sensing applications, or even food packaging.