Development of a mid-scale wind tunnel that can bridge laboratory and field experiments will be discussed. In fire and wildfire science, the problem of linking laboratory studies to the field is the main bottleneck for the development and validation of fire spread models. The laboratory offers well-controlled conditions and allows testing and validating many configurations. The field offers more realistic conditions but they are very difficult to control. The main factor that hinders applying laboratory results to the field is the wind, because the flow around any field experiment is impossible to control and extremely difficult to measure. Similarly, laboratory experiments always involve some degree of idealization, not only of the wind but also of the fuel and other environmental conditions. Our mid-scale wind tunnel has been designed to allow collecting field data for realistic fuel and environmental conditions but under well-controlled wind conditions. Hence, it represents the missing link between the idealized laboratory conditions and realistic field conditions, particularly in the context of low-intensity prescribed burns. Our wind tunnel was constructed for both laboratory and prescribed field experiments. The tunnel can be disassembled into smaller pieces allowing us to bring the tunnel to the field. The tunnel provides flow speeds up to 8 m/s with a turbulent intensity of 20% in the test section, which represents the average wind speed and its fluctuation at Silas Little Experimental Forest Research Station. The test section is large enough to test an array of shrub and litter layer structures in both laboratory and field settings. Windows enable us to measure the properties of interest such as the vegetation geometry, the flow field around vegetation, the temperature field, the flame geometry, and the fire rate of spread. For the laboratory conditions, the tunnel has a 0.2 m deep fuel bed that can be used to provide more realistic boundary conditions, soil type and moisture contents, etc. A preliminary study using simplified fuel arrays will be presented as a demonstration. The outcomes of this research will contribute to the development of physics-based predictive fire spread modeling aimed at supporting wildland fire management in a more robust way than currently available.