Area G
Part of a well-preserved wall, oriented east–west, was uncovered on the eastern side of the area (W105; Figs. 2, 3). Wall 105 in Sq L17 consisted of terre-pissé construction on stone foundations and both sides of the upper structure were coated with thick plaster. To the south of W105 was a very thick hard plaster floor (L117; thickness 0.1 m). It extended c. 3 m east to west but its width, of which c. 1.5 m were exposed, could not be determined because the southern end was in the balk. Although the floor is set parallel to the wall, its function is still enigmatic at present.
The continuation of W105 in Sq M17 was Wall 123 (width 1.5 m), whose central part had collapsed. To its south was a hard flat plaster surface (L120), similar to L117 in Sq L17. It was apparently badly damaged by fire and had signs of melting and deformation, as well as traces of charcoal.
The width of W123 and W105 corresponded to the casemate wall that had been uncovered by the former Japanese delegation. This was probably the southern side of the rectangular casemate enclosure, which suggests that the northern part of Tel ‘En Gev was indeed ‘the upper city’ and probably served as an administrative center.                 
 
A well-built Hellenistic house was unearthed in Sq M18 (Fig. 4). The outer stone walls (W138–W140) were thick and thinner walls (W145, W146) divided the building into three sections. All the floors (L144, L147, L150, L151) were coated with plaster. Combined with the results of the previous excavations, it seems that a Hellenistic town was spread over the upper city.
The three squares (I17–18, J17) opened in the western part of Area G were disturbed by a tunnel dug during the Six Day War in 1967. The only remains were two walls, oriented east–west and north–south, which apparently belonged to the Hellenistic period. Although many large stones were found around these walls, suggesting perhaps the presence of a large structure, it was impossible to draw any conclusions.
 
Area H
At the highest point of the tell, three squares (C9, CD10) were opened. A corner of a massive wall was found close to surface. Since this area is located at the northwestern corner of the mound, it probably functioned as a corner tower of the city during the Hellenistic–Early Roman periods. To its east, another wall was cut by this tower-like structure. 
Fills beneath these walls (thickness c. 1 m) were disturbed by several pits and graves. A thick plaster floor (L543, L547; Fig. 5) was uncovered below the fills throughout Sqs CD10. The floor (thickness 5–10 cm) consisted of high-quality plaster and was partly damaged by later pits. The floor makeup (thickness c. 0.4 m) was composed of two to three thin layers of white plaster alternating with brown soil and pebbles. At least three layers of plaster were discerned at the northeastern corner of Sq C10. The pottery on the floor belonged to Iron IIA (ninth century BCE, based on Modified Conventional Chronology suggested by A. Mazar [2005. The Debate over the Chronology of the Iron Age in the Southern Levant. In T.E. Levy and T. Higham, eds. The Bible and Radiocarbon Dating. London. Pp. 15–30]). The floor was not associated with any structure and hence, its purpose cannot be clearly understood. At the same time, the quality of the plaster, as well as the location of the floor and its size, may indicate that it was part of a courtyard or a hall of a large public building.
At the eastern side of Sq D10, another stratum was exposed below the plaster floor. It consisted of dark burnt red soil, two walls (W555, W556) and a tabun (L557; Fig. 6). The top part of the well-built tabun was encircled with stones. Close to the tabun, a bronze artifact and a stone bead were found, as well as a large quantity of charcoal, samples of which were collected for 14C analysis. One sample (conifer) indicated 2950±30 BP/2910 ± 30 cal BP [1 sigma: 1189–1029 BCE (68%); 2 sigma 1253–1006 cal BCE (95%)] and another sample (leguminosae) indicated 2800±30 BP/2840±30 cal BP [1 sigma: 1026–930 cal BCE (68%); 2 sigma: 1112–910 cal BCE (95 %)]. The conifer was probably remains of timber that could be used over a long period of time, whereas the leguminous branch was most likely used soon after it was cut. Hence, the destruction of the building should be assigned to the first half of the tenth century BCE. The pottery from this stratum also suggested an Early Iron Age date (eleventh–tenth centuries BCE). Although the nature of the structure cannot be defined as yet, it is possible that the plan of the upper city was largely changed around the tenth century BCE.