On January 28, 1160 King Baldwin III granted the custody and dragomanate rights of the Mi‘ilya castle to someone named John of Haifa and his heirs. A document mentions nine villages belonging to the estate, in the middle of which was the castle (Strehlke 1869: 2–3, No. 2). This is evidence that already in 1160 the castle at Mi‘ilya served as an administrative center of an estate that belonged directly to the king. On the basis of this document, several scholars assumed that the fortress was built by King Fulk of Anjou (1131–1143), whose reign was characterized by the construction of fortifications (Boas 2006:234). In February 1182, Joscelin III received the castle and everything belonging to it, save the village of Ga‘aton (Iazon) from King Baldwin IV. The castle was captured by Saladin in 1187, following the defeat of the Frankish armies at the Battle of Hittīn. With the conquest of ‘Akko in 1191, in the wake of the Third Crusade, most of the territory around ‘Akko was returned to the Franks, including Castellum Regis (Prawer 1975, II: 99). In May 1220, Beatrice (one of Joscelin’s two daughters) and her husband Otto of Hennenberg, who were then the owners of the Castellum Regis estate, sold off most of the property. The document mentions thirty-seven villages that are subordinate to the administration of the castle. In 1229, Jacob of Amigdala, the son of Agnes (Joscelin’s second daughter), demanded his inheritance rights to the estate, which the Teutonic Order had purchased from his aunt in 1220. Castellum Regis was one of the places he claimed. Montfort Castle, which was built in the late 1220’s, replaced Castellum Regis and became the region’s administrative center (Pringle 1998:30). After the Teutonic Order acquired the estate, the order also began purchasing the property of petty landowners in the village of Mi‘ilya (Strehlke 1869:120–1, No. 128). Another document from the year 1257 mentions a house and other property in Mi‘ilya that belonged to the Bishop of ‘Akko (Strehlke 1869:91–4, No. 112). The castle was presumably captured during the first siege of Monfort in 1266. Documents written by Buchard of Mount Sion from the year 1283, al-Damashqī and Marino Sanuto from the 1320s and al-Uthmani from 1364 show that the Mamluks settled in the village and castle and did not destroy them, and the castle even assumed its former status as the regional administrative center.
 
Castellum Regis (Fig. 1) is a castle on a mound, which is typologically classified as a castrum (the style of castra), also known as a “quadriburguim”. This is a square castle with four prominent corner towers (northern wall length 30.8 m, western wall length 32. 4 m, eastern wall length 33.0 m, thickness of each wall 3 m). The southern wall was completely destroyed, but in all likelihood, it was similar to the northern wall. A wall aligned north–south (thickness c. 2.3 m), whose eastern side collapsed, was built inside the castle. It divided the structure’s interior in two: three quarters of the presumed castle area are located west of the wall (22.4 m) and one quarter is east of the wall (7.8 m). A partially preserved arched passage that linked the two areas was set in the southern part of the wall. Three of the castles’ four towers were preserved; the southeastern tower was completely destroyed. 
The foundations of two towers (A, B) were discovered in the survey; parts of them are inside the vaults of houses from the Ottoman period. The western foundation of Tower B is stepped and has six stages. Each of the six steps (overall height 2.5 m) begins from a common point in the tower’s northwestern corner, whence they spread out in a fan-like manner (Fig. 2).
A Crusader arch, preserved in its entirety, was identified in the southern part of the inner western wall inside Tower A (width 1 m; Fig. 3). Based on its location, it appears to have been an entrance to a vault that was also used as an entrance to the tower, like the vault in Tower B, which was completely preserved and its width is identical to that in Tower A (1 m; Fig. 4).
A built shaft was discovered in the northern wall of the castle, close to its corner with the inner wall that divides the castle into two parts. The shaft (diam. 0.6 m) is built of finely dressed stones. Today, fifteen of its circular courses can be discerned, each course consists of three or four stones whose interior surface is circularly hewn and the exterior surface is dressed straight (Fig. 5). The upper end of the shaft is six meters high within the wall in which it is incorporated, above fill; presumably it continues down ending in a reservoir. The shaft was likely used for drawing water from the reservoir, into which rainwater would drain; it is also possible that the rainwater drained into the reservoir by way of the shaft. A shaft with an identical diameter and construction style was found in ‘Akko, and the excavator defined it as a well and said it was dug down to the groundwater (HA-ESI 110:11*–12*). Another similar well that reached down to the groundwater was discovered in excavations at the ‘Akko courthouse. The diameter of the latter well is larger (0.95 m) allowing one to descend easily and square footholds were incorporated in it one above the other, which also facilitated the descent (Hartal 1997:6; Nir 1997).
The shaft contributes to understanding the construction method of the castle. The wall was presumably erected in stages. Each time, several of its vertical courses were built, into which several courses of the shaft were incorporated; the cement was cast inside them and so on until it was completed. This method shows the structure was built according to a plan, and that the reservoir might have been hewn or built prior to the construction of the wall and shaft.
The cement in Castellum Regis is extremely strong and resistant to weathering, like at many Frankish period sites. At Castellum Regis, as at Ascalon and Caesarea, many of the stones have fallen from the walls, yet the cement is still intact.
The northern and southern firing slits in Tower B were discovered and the southern firing slit in Tower C was found completely blocked on its inside. Today, the cross-section of the firing slits appears triangular (outer height 0.85 m, inner width 1.05–1.30 m) and their ceiling is built of stone slabs set next to each other across the section and at an angle sloping from the inside to outside. A lintel is on the interior opening of the firing slits and a flat relieving arch above it (Fig. 6).
All of the castles’ slits are similar on the inside but are different on the outside. The slits on the outside of Tower A are built of two stones with drafted margins (height of each stone 0.85 m, width of opening 5–10 cm; Fig. 7). The slits in Tower B are built of four stones with drafted margins, two arranged on either side of the slit and one above the other (overall height 0.85 m; Fig. 8).
Two mason’s marks were found on the walls of the castle. Several stones are engraved with crosses; four stone have an X engraved inside a rectangle on each of the stone’s surfaces and additional less distinct marks were discerned on stones.
 
The archaeological remains indicate that the castle does indeed belong to an early type that is in keeping with the twelfth century CE, and particularly with the style that was common until the Battle of Hittīn. Other similar examples include the castle at Tel el-Safi, the castle at al-Taiyaba and even Belvoir fortress, although its plan was improved and there are two castles of this type located one inside the other. It is therefore difficult to identify construction in Castellum Regis, which can clearly be ascribed to the Teutonic Order from the thirteenth century CE.
According to the historic documentation and the archaeological remains, it seems that the castle was relatively small and mainly used for administrative purposes, although it did have some defensive importance, especially until the Battle of Hittīn in 1187 CE.
The remains indicate that the castle was completely built in one stage, such that all of the connections between the castles’ walls and its towers on the inside and out were made with bonded courses and there is no evidence of later additions. Nevertheless, the difference in the construction methods of the towers might indicate that the castle was constructed by several groups of masons.