Between the second half of the 4th century and the 2nd century BC the fortifications on the Acropolis were expanded to include the lower city. The defensive wall circuit of Hellenistic Butrint was 970 m long and embraced an area of almost 6 ha. This circuit also had at least 6 gates: Asklepieion Gate, West Gate, North Gate, Lion Gate, Lake Gate and Tower Gate. Ugolini also added some square bastions and a "bevelled bastion" to its description.
L. M. Ugolini described the "walls halfway up the hill", located over the theatre's cavea, characterised by a single curtain. He distinguished a "pseudo-pelasgic phase" in "rough polygonal" masonry on the southern side of the hill, from a "polygonal phase" with large blocks, however he did not express his opinion on their chronology. The first proper phase of construction saw the wall built all along the southern foot of the acropolis in two masonry types, described by Ugolini as "eight sided" and "six sided" polygonal. Ugolini dated it between the 5th and 4th century BC. The second phase witnessed the construction of a continuous wall around the bottom of the acropolis with the purpose of enclosing the urban centre and the lower parts of the Hellenistic city. It is built using two different masonry types, characterised by parallelepiped block masonry. However, the first's height is greater than the width, while the second has large, rectangular, and thicker blocks set in regular courses. It is dated to the 4th century BC. Finally, there is also a phase dated to the 3rd century BC by Ugolini, characterised by regular blocks of smaller dimension arranged on regular rows. This technique can be found on the stretch of wall that runs even further below the walls halfway up the hill, enclosing the entirety of the Sanctuary of Asclepius, where the double-curtain walls can be seen on both sides with their inserts in polygonal technique and a series of rectangular buttresses as reinforcement. The first circuit could be attributed to Phyrrus' construction program, with some adjustment made in the period of the koinon of the Epirotes, while the latest wall circuit might have been built when the city obtained the status of independent polis of the koinon of the Prasaiboi.
The hypotheses of Ugolini were revisioned for the first time by D. Mustilli between 1938 and 1939. The discovery of the bothros in the area between the Venetian Castle and the Acropolis Basilica led him to date the walls on the Acropolis to the 7th century BC and the expansion on the slopes to the 5th century BC.
In the 1970s N. Ceka completely revisioned Ugolini's interpretation as well, attributing new names and definitions to the wall circuit. After naming the Acropolis walls as "Butrint I" and dating them between the 7th and 6th century BC, he dated the "Butrint II" phase of the Acropolis walls, characterised by smooth trapezoidal or polygonal blocks on the outer façade, to the 4th century BC. He defined as "Butrint III" the stretch in the area from the Asclepius Gate to the Tower Gate and Lake Gate. It is dated to the middle of the 4th century BC and formed by trapezoidal blocks placed in irregular rows. Finally, he also located the "Butrint IV" walls, which he dated to the 3rd century BC.
In the 1980s G. Karaiskaj was involved again the study of the walls of Butrint, but accepted the dating proposed by Ceka.
The Hellenistic wall circuit can be traced in the area of the Lake Gate, up to the Lion Gate and the North Gate. In all these cases the building technique is pseudo-isodomic, with double curtain walls, with the exception of a section just before the North Gate. The northern side of the circuit is the well preserved. There can be seen vertical grooves, probably used for placement with a plumb bob, indents for lifting, shallow putlog holes (even though their function is still uncertain).
After the deduction as a colony in the 1st century BC, the wall circuit was extensively renovated. Karaiskaj believed that this can be seen both on the southern side of the upper circuit and on the northern side of the lower defensive circuit. Nevertheless, there is no evidence of a Roman wall circuit, even though the Hellenistic walls were maintained and repaired during the Roman era. This is particularly evident on the north side of the city, while this is less visible on the eastern and southern sides. An example is the stretch that ran from the Sanctuary of Asclepius to the Tower Gate, which was destroyed to give space to the construction of the Roman Forum, and later remodelled to be incorporated in the medieval buildings of the area.
The Roman renovations see wall made of stones with worked and bulging outer faces. They fit together without bonding materials, but on the interiori the stones are fixed in a mortar matrix including small crushed stones. Karaiskaj stated that these were build using two dry parallel walls in isodomic order and then pouring a mortar core between them, describing the technique as compromise between opus quadratum and opus caementicium. He located also the opus reticulatum and pseudo-reticolatum, which are also visible in the main Roman buildings of the city.
Giorgi E., Lepore G., “Comparing Phoinike and Butrint. Some remarks on the walls of two cities in Northern Epirus” in Caliò L. M., Gerogiannis G. M., Kopsacheili M., Fortificazioni e società nel Mediterraneo occidentale, Albania e Grecia settentrionale, Roma, 2020, pp. 153-181
Hammond N. G. L., Epirus. The geography, the ancient remains, the history and the topography of Epirus and adjacent areas, Oxford, Clarendon Press, 1967, pp. 108-109