The NIST Outdoor Structure Separation Experiments are part of the NIST Structure Separation Experiments project, which is designed to assess structure-to-structure fire spread in the wildland-urban interface. In the first phase of this project, fire exposures from small source structures (sheds) were quantified in terms of mass loss rate and peak heat release rate in the absence of wind. The performance of a target structure representing a residential façade was assessed in response to exposures from a variety of sheds (differing in construction, size, and fuel loading) and separation distances. These experiments were conducted indoors at the National Fire Research Laboratory at NIST. The knowledge gained from the indoor shed burn experiments was used to develop the outdoor shed burn experiments.This report describes a series of outdoor shed burn experiments conducted at NIST to study the effects of applied wind on thermal (radiant and convective) exposures from variety of sheds to a target structure. The results were used to determine the minimum Structure Separation Distance (SSDmin), which is defined as the shortest distance between the source (storage shed) and target structure (residence) to prevent ignition and flame spread. Experiments were conducted with combustible (wood) and noncombustible (steel) sheds, each containing a set of wood cribs and ignited to generate typical radiative and convective heat exposures to the target structure. The target structure was an assembly including an exterior wall with a window and a roof with a vented eave. Effects of shed orientation, shed size, SSD, and shed type on thermal exposure to the target structure were studied. The thermal exposure was quantified in each case by measuring the peak heat flux and temperature at the target structure. Based on target structure performances due to thermal exposure from the burning of different source structures, a SSDmin was identified. The experiments in this study demonstrated good reproducibility of experimental data. The wind had complex effects on the burning behavior of wood sheds, affecting flame lengths and convective heat transfer to the target structure. For combustible sheds, the peak heat flux measured at the target structure corresponded with the total amount of fuel was not affected by orientation (i.e., door opening facing downwind or upwind). In the case of steel sheds with the door opening facing the target structure, the applied wind had minimal or no effect on thermal exposure to the target structure. The steel sheds contained the fire effectively - thus reducing, but not eliminating, the thermal exposure to the target structure. For the scenarios evaluated, peak heat flux at the target structure could be reduced by half by changing the orientation of the door opening for noncombustible sheds. The SSDmin was identified as 10 ft for combustible and noncombustible sheds with floor area less than 26 ft2 in scenarios with a fire hardened target structure. For sheds with floor area between 26 ft2 and 64 ft², the minimum SSDmin was found to be 15 ft. Because the local winds during a WUI fire are unpredictable, SSD should be omnidirectional, i.e., the same SSD in all directions. 12 out of the 13 experiments were conducted with a fire hardened target structure. Experiment conducted with a non-fire hardened target structure to demonstrate the effectiveness of fire hardening revealed that the non-fire hardened structure ignited within 6 mins when exposed to a Very Small wood shed with total fuel of 212 kg and SSD of 10 ft. With similar thermal exposure, and SSD, the fire hardened target structure exhibited minimal thermal damage and significant ignition resistance. Acknowledging the limitations of the experimental series, implementation of the minimum separation distance identified from the experimental measurements in this study is expected to improve community resilience to WUI.