"ʿAyn el-Ǧadīda is located in the Dakhla Oasis of Upper Egypt, ca. 2.5 km northwest of the village of Ismant and ca. 5 km northwest of the ancient site of Kellis (Ismant al-Ḫarāb). The whole site is delimited to the north by the escarpment, which dramatically divides the Dakhla Oasis from the desert plateau, and to the east, south, and west by cultivated fields and patches of desert land.

The site consists of five low hills, of which mound I, centrally located, is the largest, covering an area of about one-half hectare. Three smaller mounds (II-IV) lie to the south and southwest of mound I, while mound V is located about 230 m to the northeast. Archaeological remains of mud-brick structures were identified on all five mounds, but excavation was carried out only on mound I.
The Dakhleh Oasis Project conducted a preliminary survey of the site in 1980, and the local Coptic and Islamic Inspectorate carried out three seasons of excavation between 1993 and 1995. The investigation focused on the south half on mound I and revealed a very intricate complex of rooms (several identifiable as magazines), surrounding a large, open-air central kitchen. The area showed clear evidence of a multi-phased development, with the addition of clusters of rooms built against earlier ones and extending to the outer edges of the mound. A Columbia University team, with Nicola Aravecchia as field director, excavated on the main hill from 2006 until 2008.

In the northern half of the mound, sets of interconnected rooms, sometimes opening onto apparent inner courtyards, were built against each other to form larger, roughly rectangular blocks, divided by a network of roughly perpendicular streets. Three interconnected rooms (B1-3) were excavated, i.e., two roughly square, symmetrical spaces opening onto a rectangular court (with two other–unexcavated–rooms built against the opposite side of the courtyard, for a total of five spaces). Archaeological investigation points to the identification of this sector as a residential area. A remarkably large structure was also surveyed along the north end of mound I, consisting of two rectangular rooms located at the center of a wide, rectangular court (ca. 16 m north-south by 12 m east-west). The complex was likely a pigeon tower, a common feature in the oasis landscape.

The central part of the hill reflects a more irregular arrangement, which might be the result of a less planned, multi-phased rearrangement of space. A church complex is located in the middle of the mound. The church (B5) consists of a one-nave room, originally barrel-vaulted; it is oriented to the east and ends with a round apse and L-shaped pastophorion, both added at a later time against the east wall. Mud-brick benches run along the north, west, and south walls of this room, which was originally accessible from the north through two doorways, one near the northwest corner and a large opening in the middle of the north side. These connected the nave with a large, rectangular hall (A46), a space originally covered by a barrel-vaulted roof and with benches built along the north, east, and south walls. Against the south side of the central passageway is a stepped mud-brick podium, accessible only from the church, that once granted people in both rooms the possibility to see and hear the person standing on the podium. At some point, this passageway was closed and the podium eliminated. The gathering hall opens to the north onto a smaller, barrel-vaulted rectangular space (B6), which was used, at least in its latest occupational phase, for the preparation of food; indeed, a hearth with traces of ash and charcoal was found against a low, rectangular platform along the north wall. Also, imprints of jars are visible on a raised platform against the east wall and along the south wall of the room at floor level. Some graffiti can be seen on the west and north walls, including two inscriptions –one in Greek and the other in Coptic– and some drawings such as boats and a bird. This anteroom is accessed from a long corridor (B7) running east-west to the north of the gathering hall. This space, which seems to have been roofless, ends to the east with a doorway that is the only entrance into the church complex from the outside. A narrow, vaulted passageway connects the anteroom with another space to the north (B9). A mud-brick recessed feature, possibly used as a cupboard in antiquity, is set in the south wall of this room, by its southeast corner. It is possible that this space, which did not open onto any other room besides anteroom B6 to the south, was used as a magazine for the latter. A doorway located in the northeast corner of the anteroom opens onto a well-preserved staircase (B8), which led to the roof of a kitchen (B10).

A long street (B12) runs from north to south along the east side of the complex. It consists of three different sectors, which were part of a longer street running from north to south across the hill. A small open-air area (B14-15), with evidence of several bread ovens, once opened onto this street near the main entrance of the complex. Another passageway (B11), which was originally barrel-vaulted, runs east-west along the south wall of the church and intersects the north-south street at its east end, where an open courtyard (B13) is located. Two clay and mud-brick rectangular features, possibly associated with the feeding of animals, are built against the south wall of this crossroads.

A considerable amount of data proves the existence of different phases of construction within the church complex, including the south wall of the church (irregularly laid out and clearly built in different phases), the semi-circular apse with L-shaped pastophorion (added at a later stage), and the mud-brick wall blocking the podium. Furthermore, foundation walls in the church and the hall to the north belong either to previous buildings or to earlier construction phases of the complex. The available evidence suggests that three smaller rooms originally stood in the area later occupied by the church, the gathering hall, and the anteroom. The walls of these early structures were either razed or incorporated within the walls of the church complex, which was extended to the west and northwest, but also to the east with the addition of the apse and pastophorion.

The closest parallel to the plan of the church of ʿAyn el-Ǧadīda is the Small East Church of Kellis. The church, excavated by Gillian Bowen, has comparable dimensions and construction material and has a significantly similar–although not identical–layout, with a large rectangular space opening to the south into an apsidal room through two doorways, a smaller one to the west and a wider passage in the middle. The dating, which is based mainly on coins, is consistent with that of the church complex at ʿAyn el-Ǧadīda.

About 25 m to the northwest of the church, and lying along the western edge of mound I, near the cultivated fields, is a large, rectangular complex of eight rooms (B17-24). It measures 18.50 m north-south by 7.10 m east-west and consists of an extensive courtyard with niched walls to the west and east (only the niches of the east wall are partially preserved); to the north of the courtyard (which was later partitioned into different rooms) is a set of two very small rooms (B20-21), opening onto each other, flanked by two rectangular side-rooms (B22-23). The complex can be identified as originally a small-scale mud-brick temple, converted into a ceramic workshop in Late Antiquity. Indeed, several pieces of unfired pottery were collected during the excavations, together with large lumps of clay and the extensive remains of clay (or clay and stone) bins embedded in the floor. To the east of the complex is a rectangular space (B4) that contained a large quantity of ash, charcoal, organic material, broken objects, and pottery sherds. This evidence suggests that the space was used, at least in its latest phase, as a domestic midden.

All buildings excavated or surveyed at ʿAyn el-Ǧadīda have walls built of sun-dried mud-bricks, rich in organic inclusions. Their dimensions fit the standard measures of Roman samples, which were generally adopted in Coptic architecture. The bricks used to build the vaulted ceilings are of a larger size. Stone was rarely used at ʿAyn el-Ǧadīda, mostly for the lintels of doorways. No wooden feature was found in situ; however, wood was certainly used as a building material, especially for doorways and shelves.

The material evidence from ʿAyn el-Ǧadīda includes ceramics, lamps, ostraka, coins, wood and metal (especially bronze) objects, dull glass jewels, and items made of vegetal fibers. Pottery is the largest category of finds retrieved on site, including thousands of fragments and a few complete or almost complete vessels. The ceramic corpus can be assigned to a period spanning from the early fourth to the beginning of the fifth century CE, with very few exceptions. The readable coins from the site largely confirm this chronology; most are from the fourth century CE, and mainly from the first half of the century. Significant evidence for an earlier chronology than the fourth-early fifth century CE was provided by the discovery of the mud-brick temple near the west edge of mound I. This complex testifies to the existence of the settlement since probably the second century CE.

The archaeological record has not provided any evidence suggesting episodes of violent destruction or abrupt abandonment. No objects of significant value were found in the rooms that were excavated.

ʿAyn el-Ǧadīda was preliminarily identified, since the time of the first survey and of the SCA excavations in the mid-1990s, as a rural village or monastic settlement. One reason stimulating the latter interpretation is the clustered configuration of the south part of the main hill, with a large kitchen centrally placed. This was seen as typical of a spatial arrangement meant to satisfy the needs of people who lived a communal life, rather than being divided into standard families. But the site as a whole was certainly not monastic. It may have been an epoikion, that is to say, an agricultural settlement with wage or tenant labor, quite possibly seasonal, dependent on Kellis–of a type largely attested in documentary sources but never fully documented by excavation."

- Nicola Aravecchia