A massive square structure (c. 12 × 12 m; Figs. 3, 4) revealed in the northern part of the excavation area consisted of large basalt stones with drafted margins on their exterior face, characteristic of public buildings and fortifications in the Hellenistic and Early Roman periods. The walls of the building were erected in two close, well-planned phases: in the first phase, the interior walls (W109–W111, W123; width c. 1.5 m) were apparently built of roughly hewn basalt stones with their flat, worked side facing out. In the second phase, the ‘casing’ walls (W122, W125, W138, W143; Fig. 5) were constructed, similar in width to the interior walls. Both faces of the outer corner stones were dressed. The interior walls were built mainly of stretchers (Fig. 6) with smaller stones arranged lengthwise above them. The outer wall was built of headers-and-stretchers, whereby one header was placed between every two stretchers. The walls of the two phases together are more than 3 m wide.
A room (1) was excavated in the center of the structure, exposing white plaster that was preserved on most of the area of the floor and on some of the walls (L126). In the center of the room were three column bases that were built on a foundation oriented northeast–southwest (Figs. 7, 8) that divided the room into two equal spaces. Stone or wooden columns were probably placed on these bases. A round plastered depression (L127; depth c. 0.5 m) exposed in the southern part of the room may have been used as an installation for a jar or to collect liquids from the many jars whose fragments were found there. A portion of the floor was removed (L133, L134) and the excavation continued to bedrock.
The interior of Room 1 was found filled with large building stones that caused extensive damage to the plaster when they collapsed. The stones survived as steps of sorts, and they were probably the remains of an entrance platform or a staircase. The collapse may have been the result of an earthquake, as evident in the northern wall (W125; Fig. 9).
The room was entered from the southeast, by way of an entrance platform or steps (L129) leading from a corridor and a small room (L113) that was blocked at its southeastern end by large square basalt blocks (W115). A doorway (Fig. 10) that had been intentionally blocked with stones and earth was discovered in the southern part of the room, in the center of W110. After clearing the blockage, a passage (L142) was discovered; it was partially excavated and led to another space that remained unexcavated. An entrance consisting of a doorjamb and threshold made of basalt stones with drafted margins (L140) was revealed in the western part of the structure, outside the line of walls. The southern side of the entrance was adjoined by a staircase with basalt fieldstones incorporated in its base; however, this might have been the level of the natural bedrock (L141). A fieldstone wall (W139) founded on bedrock was built from the entry doorjamb and along the wall of the structure (W122). The excavation in this area was not completed and it was not possible to define the complex. Another smaller entrance threshold was discovered in the northeastern part of the structure, but here too the excavation was not finished and the connection to the building between the two entrances and this threshold is unclear.
Remains of gray plaster were exposed on top of the foundation stones and on bedrock in the excavation conducted outside and north of the structure (L137) and in the outer part of W125. The plaster consisted of several layers and covered the bottom of W125 until the course of stones with dressed margins and was probably intended to prevent moisture from penetrating into the lower level of the structure. The color and texture of the plaster was different than the plaster on the floor of Room 1. A wall built of large fieldstones and square stones (W120) was exposed southwest of L137; it abutted W125 and was perpendicular to it. The wall was founded on top of the basalt bedrock and the plaster remains; thus, it was later than the building but its dimensions and function are unknown at this stage.
The excavation revealed a massive state-owned building constructed in two stages. It was strategically located above Nahal Roqad, and had an excellent command of the surrounding region. In a later phase, a wall or another complex was built to the north, the date and purpose of which were not ascertained. The building style is in keeping with the Hellenistic and Early Roman periods. Fragments of pottery vessels and lamps that were collected in the excavation indicate that the structure was used in the first half of the second century BCE, and following a hiatus, was used again from the late first until the late second centuries CE.
The importance of the excavation lies in its discovery of a unique structure in this part of ​​the Golan—its location, shape, size, massiveness and date of use.
The structure may be related to the governmental regional defense system erected in the region after the Battle of Panium in the early second century BCE. Further study is likely to contribute to our understanding of the eastern territory of Susita ​​(Hippos) and perhaps, to our understanding of the relations between the cities of the Decapolis— Hippos, and Gedera, to its south (Maoz 2013:78–84; Eisenberg 2016). The relation between this site and other sites in its vicinity is unclear. It is not known whether routine everyday ties were maintained with the stronghold or whether it was an independent governmental/military facility in the area.