The structure was built on a moderate slope, directly upon a layer of sterile alluvial earth. Only four walls, built of medium to large sized fieldstones, were discovered. The delineation of the walls enabled the identification of four separate rooms (Loci 105, 107, 110, and 112). Due to the limited scope of the excavation, the rooms were only partially excavated and thus, little can be said about their function or complete size (Fig. 3). 
The westernmost rooms of the building, Rooms 107 and 112, were separated by a single wall (W3). Wall 1 was the northern border of these two rooms; it was perpendicular to Wall 4. At the southern extent of W3, no continuation of the wall was preserved and it seems to have ended abruptly. No clear intrusion was noted and thus, the gap may indicate a possible entranceway, connecting between the rooms. A beaten-earth floor (L107) was discovered in Room 107, abutting the base of W3. The floor was overlain with numerous flat-laying potsherds (Fig. 4). Room 107 may have served as the structure’s central courtyard, around which other rooms were built.
Room 112, to the north of Room 107, had also a beaten-earth floor (L112), which abutted the base of W3; near the mid section of the wall, a rectangular shaped flat topped stone was uncovered. The stone was embedded into the floor surface and may have been used as a pillar base providing support for an assumed roof.
The easternmost rooms of the building, 105 and 110, are separated from the western rooms by W1 and W2. A partition wall (W4), oriented northeast-southwest, separated between the two rooms. Room 105, the least excavated of the rooms, was enclosed by its western (W2) and northern (W4) walls. A small gap between the walls would have allowed the entrance into the room from the adjoining Room 107. The floor in Room 105 was composed of packed beaten earth (L105).
Room 110, to the north of W4, was partially uncovered. It was nearly completely filled with a layer of jumbled stone collapse. Upon removal of the stones, the southern face of W1 was uncovered, preserved 3–4 courses high (1 m). A beaten-earth floor (L110) was uncovered at the base of the wall. The floor elevation was noticeably lower (0.7 m deep) than in all other rooms of the building. A partial stone pavement, composed of a few flat topped stone slabs, was discovered in the central area of the Floor 110. A restorable store jar was found above the stone pavement.
Both potsherds and flint tools, indicative of the IBA period, were recovered from all areas of the excavation. The ceramic assemblage was dominated by store-jar fragments, featuring flat bases and triangulated rims. These vessels often show vertical incised or applied thumb imprinted decorations around the neck.
 
The current excavation was the first to identify and sound protohistoric remains dating to the IBA period within Moshav Tarum. Only a few sites dating to this period are known in the immediate area. Salvage excavations, c. 2 km to the northeast of Moshav Tarum, along the southern edge of Road 44, were carried out in the winter of 2011 (HA-ESI 124), revealing numerous built pebble and cobble surfaces in conjunction with terrace walls. The potsherds within the stone surfaces dated to the IBA period and it seems that during this period the area was used for various undetermined agricultural purposes. Remains of a multi-roomed farmhouse, dating to the IBA period were excavated in 2011 at Nevē Shalom (Permit No. A-6323), located 4 km to the north of Tarum, near the Latrun Junction.
Virtually no significant stone collapse was uncovered at Tarum, indicating that the walls probably served as foundations for a much larger construction of mud bricks, although no clear mud-brick material was discerned. Wall 1 was found in the best state of preservation among the walls, proving that a great deal of preparation and effort was taken into building the structure. The significantly lower floor level of Room 110, as compared to the higher floors in Rooms 107 and 112, suggests that the natural topographic slope was intentionally altered and terraced prior to construction. The earth was cut and sectioned, creating two-terraced platforms, and then small and medium-sized stones were placed along the section to serve as a foundation for the upper larger stone course of W1. The flat topped stone within the floor makeup of Room 112 is assumed to be a pillar base, indicating that the room was apparently roofed.
The architectural remains uncovered are incomplete and only reflect a portion of a large building that was part of an assumingly larger settlement, whose extent is not known. Due to the limited scope of the excavation, little can be said about the nature of the structure and the uncovered rooms.