Deep stratigraphic constrains below the Indonesian Lusi mud eruption are currently lacking due to the absence of deep wells and good quality seismic data. A collection of carbonate clasts has been sampled from the Lusi site, active since its birth in 2006. These specimens are part of a large variety of lithotypes erupted from the main crater. The carbonates analysed comprise scleractinian coral and bivalve shell fragments, probably shallow- water in origin, and clasts consisting of planktonic foraminifera-bearing mudstone, from pelagic deposits. Selected rocks were analysed using planktonic foraminifera and 87Sr/86Sr dating with the aim to constrain their age and to improve the understanding of the, so far unknown, sequence of limestone deposits inferred at this site. Based on biostratigraphy using planktonic foraminifera, one group of samples reveal to belong to the Planktonic Foraminifera Zone M5b, with an age comprised between 16.29 and 15.10 Ma (Miocene, Latest Burdighalian to Langhian). The Sr isotope-based ages of clasts analysed for 87Sr/86Sr cover a larger time window spanning from Pliocene (Zanclean and Piacenzian), Miocene (Messinian) down to Eocene (Priabonian). The Pliocene and Messinian ages are unreasonably young from what is known of the local geology and one sample provided an 87Sr/86Sr age that is ∼8 My younger compared to that obtained from the planktonic foraminifera assemblage occurring in that sample. The discrepancy suggests that this and the other unduly young samples have possibly been contaminated by geological materials with radiogenic Sr isotope composition. The minimum age of 37.18 Ma obtained by 87Sr/86Sr from a well-preserved oyster shell indicates that some of the clasts can be attributed to the deep seated Ngimbang Formation. The dating has been combined with the interpretation of seismic profiles to investigate the stratigraphy of inferred carbonates below Lusi and the PRG-1 well located some kilometres to the north east. PRG-1 borehole data are also integrated and discussed in this study. The obtained results reveal multiple implications: a) the Tuban and Kujung Formations are overlapping at Lusi site; b) the Lusi feeder conduit brecciated and mobilized to the surface carbonate lithologies buried as deep as possibly ∼3.8 km as well as even older and deeper seated lithotypes from the Ngimbang Formation; c) since the deeper carbonate samples erupted in 2006 belong to the typically not overpressured Kujung Formation, an additional overpressure generated from deeper units (Ngimbang Formation) would be required to force these fragments to surface.