The Civil Survey of 1654 to 1656 recorded it bluntly as 'a small pile of stone out of repair', which is about as dismissive an official description as a building can receive. What survives today bears that out: a two-storey structure measuring roughly eight metres by less than three metres internally, built from local shale without a single piece of dressed stone, and showing no evidence of a fireplace or a staircase. It is an austere, pared-back thing, and the closer you look at it, the more questions it raises about how it was actually used.

The earliest firm reference to the site dates from the early seventeenth century, when a castle here, possibly an earlier building on the same ground, was associated with David Synnott, who by the time of the Civil Survey held 180 acres at Brownscastle. The structure sits on a rock outcrop above the east-west Corrock river, about sixty metres to the north, in a steep-sided valley whose southern escarpment is cut by a natural defile immediately to the west. That defile carries a road to Taghmon, roughly two kilometres to the south, but an older route winds up the escarpment from the northeast and passes directly beside the ruins before turning southeast. The building's position relative to these routes suggests it was intended to watch or control movement through the valley, though its modest scale sits oddly with any grand defensive ambition. The west doorway opens not into open ground but into a D-shaped walled enclosure, its masonry still standing around one and a half metres high in places, with four small window lights and a wide opening to the northwest that may be the result of later damage rather than original design.

The internal arrangement is spare to the point of peculiarity. The ground floor has slit windows set into both the long walls and the end walls, and the ceiling joists were laid directly into the masonry. Low arches, about a metre deep, are set against the inner faces of the north and south walls, spanning the width of the building. The first floor has narrow lights on all four walls, one of them paired with a horizontal slit below. With no trace of a hearth and no sign of how anyone moved between floors, the building resists easy categorisation as a comfortable residence. The Down Survey parish map of 1656 to 1658 shows the structure was already being recorded as a feature of note, even in its dilapidated state, which at least confirms it had been there long enough to acquire a certain local permanence.