Recent developments in the radar application observation behind corners, i.e., of surveillance volumes that are off the line of sight to the radar has been reviewed and is reported here. This report includes both work performed at FOI and by external groups. Considerably less amount of work has been published about this application than that about another radar use in urban scenarios, viz., through-the-wall observation. The published accounts described in this report are experimental and theoretical. In the experimental work only relatively simple scenarios and short distance surveillance have been studied. This has been done using general purpose measurement equipment, not specifically tailored for the application. Frequency stepping has been the method to cover the wide bandwidths used, with center frequencies of about 10 GHz. In the scenarios with targets specifically located behind corners, the measurements have used quasi-monostatic horn antenna arrangement. Detections have been made of moving reference reflector and human targets. In the latter case Doppler detections have been made both of translational movement (walking, running) and of micro-movement of body parts including breathing. As expected from previously performed calculations, the strongest detections are obtained from propagation paths to the targets involving one or several specular reflections in adjacent walls, but also propagation paths to the target via corner diffraction have been identified. This latter case is judged to be of secondary importance, due to the weak signal levels involved. Much of the emphasis has been on obtaining some understanding of multipath propagation phenomenology in the urban scenario. Specific measurements have been dedicated to studying scattering from walls with roughness. It has been realized that multipath propagation could be used to the observer's advantage since it offers richer, multi-aspect scattering information about targets, which potentially may be used, e.g., for target localization with tracking and rudimentary imaging. These aspects have been investigated in work that has used both data collected from measurements and simulated data for stipulated target/environment scenarios. In the latter case, adaptive (cognitive) signal choice has been introduced. Performance improvements have been demonstrated when introducing such signal processing thriving on multipath propagation. Multistatic sensor arrangements (MIMO) have been studied in simulations to improve on, e.g., target positioning. This application is still at an exploratory stage, with considerable development being necessary toward demonstrated operational standard. Considerable further measurements are needed to develop a base of data from realistic settings for simulations and algorithm testing. Development work is required to produce equipment tailored for the application, including the use of array antennas and signal choice for coherent, real-time processing, with power generation adapted for the situation at hand. The multipath situation encountered in an urban environment has to be further examined and understood, to its more efficient utilization for performance improvement as regards target localization, etc. Continued use must be made of advancements in signal processing, e.g., with respect to adaptive and cognitive systems.