A promontory fort works on a simple principle: the sea does most of the defensive work on the exposed sides, and human labour concentrates on the narrow neck of land connecting the headland to the mainland. Here, that labour has worn down to something subtle. The bank barely reaches 0.4 metres in height today, yet the logic of the place, an oval peninsula measuring roughly 185 metres north to south and 250 metres east to west, is still legible in the landscape.

The antiquarian T. J. Westropp described the site in detail in 1914, naming the headland Gubadoon, and it appeared on the 1920 Ordnance Survey edition as an entrenchment. What he recorded and what survives today do not entirely match. The section through the defences that Westropp documented has since disappeared, and the interior earthworks, visible as faint banks from the air, are partly obliterated on the ground by the relatively recent cutting of scraws, the surface layer of turf and root that was sometimes used as makeshift field fencing. Near the western end of the neck bank, a 2-metre gap flanked by upright slabs is likely the original entrance, and a faint rise in the ground suggests a causeway once allowed passage across the fosse at this point. The fosse itself is shallow, reading now mainly as a band of darker vegetation about 5 metres wide, deepening to around 1 metre only toward the western end where some stone facing is still visible. The cliffs of Dooega are visible to the north, and the ground drops steeply from that direction toward the fort, giving whoever once occupied this headland a commanding view across a landscape of mountainous bog and thin sheep pasture that looks much the same today.