Excavations were undertaken in the rectangular entrance (A), 5 × 4 m, of the building situated in the south-western part of the site, in the immediate vicinity of the south gate in the curtain wall. The building had a rectangular plan on a north-east/south-west alignment, measuring circa 11 × 12 m and constituted by at least three rooms (A, B and C). The structure, whose north wall was not found, could originally have reached a length of circa 20 m. It can be interpreted as the palatio novo, mentioned in a document of 1339, probably constructed between the end of the 13th-beginning of the 14th century. The building had two levels: the ground floor (perhaps just one room), was delimited to the north-east by the rock face, and the upper floor was built to the north-east on the bedrock and projected to the south-west. The only entrance identified, situated on the ground floor in the south-eastern side, had a double wooden door, closed by a double system of horizontal poles which ran inside the walls and by an iron lock. The excavation revealed the presence of a staircase up against the south-western wall, comprising a masonry built base and wooden steps. It was partially rebuilt between the end of the 15th-begionning of the 16th century. The stratigraphy showed that, at the beginning of the 16th century, following circa a century of abandonment, the building was restructured and transformed into a guardhouse for soldiers sent by the Florentine government to control the Val Bisenzio for fear that this was the route chosen by the Spanish troops marching on Prato and Florence.

The large space forming the ground floor was divided into three distinct rooms (A, B and C) and the floor newly paved with shale slabs and tile/brick fragments laid flat. The upper floor was also probably remade and presented a paving of mezzane tiles. The roof was constituted by imbrices and tiles and probably had a single slope, north-east to south-west. The hearth was positioned at the centre of the space, as attested by traces of thermo-transformation and a thin layer of ash and charcoal. It is difficult to date and could have been used both by the soldiers and later, occasional occupants. The structure seems to have been definitively abandoned between the end of the 16th century and the beginning of the 17th century, when the upper floor, onto which part of the roof had already fallen, collapsed. The smaller number of imbrices and tiles with respect to the mezzane tiles may indicate a partial recovery of this material before the collapse of the roof itself. The large amount of pottery including intact or reconstructable vases, polychrome graffita from the Florentine and Emilian areas and majolica from Montelupo datable to between the end of the 15th century and the first half of the 16th century, associated with course ware testi, may attest the presence of a refectory and kitchen on the upper floor.