Brown Mountain is a flat-topped ridge on the edge of Linville Gorge, in the Pisgah National Forest. The usual watches are the overlook on N.C. Highway 181, about mile marker 20 north of Morganton; Wiseman’s View; and the Lost Cove overlook on the Blue Ridge Parkway. Witnesses describe small lights, often red, white, or yellow, that rise, hover, fade, and sometimes seem to split. Those are reports of how the lights look. The descriptions do not agree with one another from one watcher to the next.

The Highway 181 sign tells visitors the lights have been seen for hundreds of years, and it gives the Cherokee battle story. It also tells of a woman searching with a torch for a fiancé who never came, and of a murdered young woman whose light still shows. A separate telling, heard at the same overlook, is a lantern carried by an enslaved man searching for his master. That lantern story is also told of Cole Mountain in West Virginia, with a different name on the hunter. George Rogers Mansfield, who investigated for the Geological Survey, found that many local people had not heard of strange lights until about 1910. Joseph Loven, who lived near the old hotel that faced the ridge, said he had noticed a light in 1897 and that public talk began when a neighbor, C. E. Gregory, pressed the subject around 1910. Mansfield wrote that before electric lights, a favorably placed lamp might have flared in the mountain air, and that such sights would have been too few to make a local wonder. The first known print account is the Charlotte Daily Observer, 13 September 1913.

In October 1913, at the request of Representative E. Y. Webb, Survey geologist D. B. Sterrett spent a few days on the ridge. He matched lights seen from Loven’s, about six miles beyond Brown Mountain, to the headlights of westbound Southern Railway locomotives and to the train schedule. He called locomotive headlights the cause, beyond doubt.

In July 1916 a flood on the Catawba washed out bridges and track around Morganton. For several weeks no train came within forty or fifty miles of Rattlesnake Knob, and the lights were still reported. Mansfield used that report for one limited conclusion: the lights could not be locomotives alone. Automobiles were already on the valley roads. The flood does not show a valley with no headlights in it.

A 1919 Weather Bureau note compared the lights to the Andes light, an electrical display, without a visit to the mountain. Mansfield set that class of explanation aside.

In 1922 Mansfield spent two weeks in the area with a plane table and a telescopic alidade, working from Loven’s and from other points, including the mountain itself. Of the lights he could map, about 47 percent were automobile headlights, 33 percent locomotive headlights, 10 percent stationary lights, and 10 percent brush fires. He wrote that the lights were clearly not of unusual nature or origin. At the distances involved, a car headlight and a locomotive headlight look alike. Southern Railway headlights of that period, fitted with a 16-inch reflector, were rated around 600,000 candlepower, bright enough to see for tens of miles. Dust and mist after rain scatter the beam and give it the yellow and red tints watchers describe, and a thicker mist puts the light out. He did not assign a share of the lights to marsh gas.

Daniel Caton, a physicist at Appalachian State University, has watched the ridge since the mid-1980s. With observatory engineer Lee Hawkins he runs two time-lapse cameras: one at Gingercake, dusk to dawn, with Brown Mountain in front of the lights of Lenoir, and one looking up Linville Gorge from the south, midnight to 5 a.m. He has said that almost everything visitors call a Brown Mountain light, once it is looked at with binoculars, is a campfire, a lantern, a headlight, a tower, or an aircraft. On the night of 16–17 July 2016 both cameras recorded a light over the ridge that came on and went out four times in a few minutes. Caton said he had ruled out the likely man-made sources for that clip and had no explanation left for it.