A principal goal in sensory neuroscience is to understand how properties of our environment are reflected in neural activity patterns. Recent advances in computational modeling provide increasingly accurate predictions of how neural populations across the brain respond to complex naturalistic stimuli. The employed computational models, referred to as encoding models, explicitly transform complex stimuli into observed neural responses. This rapidly developing field is becoming increasingly important in sensory neuroscience as it provides detailed insights into the functional organization of neural representations. The present work starts by discussing the theoretical underpinnings of encoding models. Next, various applications of encoding models are reviewed. Finally, potential research directions that may shape future work in this area of research are described.