<strong>Original citation:</strong> SE Palmer, T Ghose (2008). Extremal edges: a powerful cue to depth perception and figure-ground organization. Psychological Science, 19.1: 77-84. <strong>Target of replication:</strong> Extremal edges (EE) occur when our line of sight runs parallel to some portion of a smooth convex surface. They are thought to be an essential aspect of scene analysis and a strong cue to figure-ground organization. The experiment we are replicating focuses on establishing the significance of extremal edges as depth cues. In the original study, participants chose which of a pair of adjacent surfaces was closer and more figural. The authors of this study reported that surfaces with an EE along the shared contour between the two surfaces were judged as closer than equally convex surfaces with an EE orthogonal to the shared contour. We are replicating their second study, which showed that the effect persisted even when the only indication of edge type was given by texture. <strong>A priori replication criteria:</strong> We decided that the key result of this study was the most stringent test of the theory: the finding that surfaces with a single EE along the shared contour were perceived as closer than surfaces with two EE’s orthogonal to the shared contour. <strong>Materials, Data, and Report Study materials</strong> can be found in the materials component of this project. Raw data and analyzed data are contained in a single file (EE_allData_NV.xlsx), which is located in the dataset node. The same node also includes a detailed description of the data in dataset_log.rtf. <strong>Conclusions.</strong> Our data fell short of supporting the original study’s claim that extremal edges are used as depth cues in scene processing. In our data, surfaces with two extremal edges parallel to the shared contour were chosen as closer than nearly all other textures, paralleling the original study. However, we were unsuccessful in replicating the finding that surfaces with a single extremal edge along the shared contour are perceived as closer than convex surfaces with two extremal edges orthogonal to the contour. In the original study, this comparison was the most stringent test of the role of extremal edges in figure-ground organization, as distinct from convexity. Consequently, it was the target of our replication. Analyses of individual differences help clarify our mixed success in replicating this study. Our participants were members of the Cambridge, MA community, a more diverse sample than the UC Berkeley undergraduates that participated in the original study. Though the broad age range of our subject pool did not seem to affect the results, there was nevertheless massive variation in responses between subjects. Thus, the heterogeneity of our subject pool may have weakened the results.