When the federal agencies established policies in the late 1960s and early 1970s to allow the use of natural fires in wilderness, they launched a natural fire management experiment in a handful of wilderness areas. As a result, wildland fire has played more of its natural role in wilderness than anywhere else. Much of what we understand about fire ecology comes from observations of natural fires in several wilderness areas that have been allowed to burn under a wide range of physical and biological conditions since the 1970s. Wilderness fires have provided valuable datasets for improving fire history methods and understanding of the drivers of fire. Inside some wilderness areas, enough data have accumulated from multiple repeated fires at natural fire intervals to see how forests respond to fire. As a result of the wilderness fire management experiment we can better anticipate the consequences of reintroducing fire and whether restoration with natural fire might be feasible. The experience of allowing fires to burn in wilderness has also contributed to social science knowledge. Studies have examined how public support for the use of fire in wilderness can change over time. Studies of the institutional factors that influence the use of fire in wilderness have pointed to difficulties with implementing wilderness fire policy, as well as the importance of belief and commitment of an individual line officer in overcoming obstacles to carry out a wilderness fire program. Future trends in climate and land use will exacerbate current challenges for wilderness fire management programs, and making the decision to allow fire to burn in wilderness will increasingly demand scientific information and will likely require an even more firm belief in the value of natural fire.