The tissue engineering approach for repairing osteochondral (OC) defects involves the fabrication of a biological tissue scaffold that mimics the physiological properties of natural OC tissue (e.g., the gradient transition between the cartilage surface and the subchondral bone). The OC tissue scaffolds described in many research studies exhibit a discrete gradient (e.g., a biphasic or tri/multiphasic structure) or a continuous gradient to mimic OC tissue attributes such as biochemical composition, structure, and mechanical properties. One advantage of a continuous gradient scaffold over biphasic or tri/multiphasic tissue scaffolds is that it more closely mimics natural OC tissue since there is no distinct interface between each layer. Although research studies to this point have yielded good results related to OC regeneration with tissue scaffolds, differences between engineered scaffolds and natural OC tissue remain; due to these differences, current clinical therapies to repair OC defects with engineered scaffolds have not been successful. This paper provides an overview of both discrete and continuous gradient OC tissue scaffolds in terms of cell type, scaffold material, microscale structure, mechanical properties, fabrication methods, and scaffold stimuli. Fabrication of gradient scaffolds with three-dimensional (3D) printing is given special emphasis due to its ability to accurately control scaffold pore geometry. Moreover, the application of computational modeling in OC tissue engineering is considered; for example, efforts to optimize the scaffold structure, mechanical properties, and physical stimuli generated within the scaffold-bioreactor system to predict tissue regeneration are considered. Finally, challenges associated with the repair of OC defects and recommendations for future directions in OC tissue regeneration are proposed.