Understanding the links between subducting slabs and upper-plate dynamics is a longstanding goal in the field of tectonics. Upper-plate dynamics are thought to result from processes that drive forearc growth or recession, which are important for understanding the evolution of the continental crust, recycling of fluids, petrogenesis of arc magmatism and earthquake nucleation and propagation. However, where converging plates first meet at subduction zones are marine environments, limiting data collection density and resolvability. Costa Rica, a relatively well-studied convergent margin, is thought to be in a state of recession, where erosion and removal of the underside of the forearc dominates and causes forearc retreat. Recently collected 3D seismic reflection data offshore the Osa Peninsula, Costa Rica, yields a rare opportunity to assess this model and to characterize the internal structure of the outer forearc and underthrusting plate of a subduction zone. This dissertation investigates the Pleistocene to present evolution of the outer forearc and the in situ properties of the megathrust. Chapter 2 utilizes sequence stratigraphy methods to extract Pleistocene vertical motions across the slope and shelf and finds that the timing and rates of these motions conflict with the commonly accepted model of continuous subsidence tied to basal erosion. Importantly, Chapter 2 demonstrates that the outermost forearc has not migrated landward during the Pleistocene as previously proposed. Chapter 3 is a study of the geometry and physical properties of the megathrust, showing for the first time in situ corrugations and abandonment and reestablishment of the megathrust up-section, which transiently increases roughness and may inhibit earthquake propagation. Chapter 4 quantifies the in situ material exchange between the plates across the outermost 10 km of the forearc and discovers a state of equilibrium where accretive and erosive processes are variable in space but when integrated are approximately balanced. In this accounting, sediment subduction is the dominant process. When scaled up to the length of the forearc, basal erosion cannot wholly account for the observed record of slope subsidence and trench embayment. Collectively, these studies characterize in 3D a margin with characteristics thought to be tied to net forearc erosion but are likely tied to subducting plate dynamic changes and shortening across the forearc. Because this is the first 3D seismic reflection dataset imaging a non-accretionary convergent margin, these results may be indicative of processes ongoing at other margins.