Policing tactics that are proactive, focused on small places or groups of people in small places, and tailor specific solutions to problems using careful analysis of local conditions seem to be effective at reducing violent crime. But which tactics are most effective when applied at hot spots remains unknown. This article documents the design and implementation of a randomized controlled field experiment to test three policing tactics applied to small, high-crime places: 1) foot patrol, 2) problem-oriented policing, and 3) offender-focused policing. A total of 81 experimental places were identified from the highest violent crime areas in Philadelphia (27 areas were judged amenable to each policing tactic). Within each group of 27 areas, 20 places were randomly assigned to receive treatment and 7 places acted as controls. Offender-focused sites experienced a 42 percent reduction in all violent crime and a 50 percent reduction in violent felonies compared with their control places. Problem-oriented policing and foot patrol did not significantly reduce violent crime or violent felonies. Potential explanations of these findings are discussed in the contexts of dosage, implementation, and hot spot stability over time.