The reliability of biostratigraphic correlations in neritic carbonate platforms is often questioned because the benthic fauna on which biozonation is based are particularly sensitive to environmental change. It is crucial to know whether a population change corresponds strictly to a facies change. Conversely, there arise the questions of determining how populations are renewed over time and how new species appear even if facies associations remain unchanged. This is the case with the Gavrovo-Tripolitza zone of Greece, an isolated shallow carbonate platform surrounded by two oceanic domains (Pindos-Olonos Zone and Ionian Zone). The absence or scarcity of faunas generally used in Upper Cretaceous biostratigraphy has led to the use of local biozonation instead, based on faunas endemic to Adriatico-Aegean platforms. The final two biozones based on Rhapydioninidae foraminifera are: - CsB6 (Upper Campanian-Lower Maastrichtian): the "Murciella biozone" is the total range zone of all Rhaphydionininae except for Rhapydionina liburnica; - CsB7 (Upper Maastrichtian): the total range zone of R. Liburnica. The purpose of this paper is to test the biostratigraphic value of the benthic foraminifera by comparing the distribution of the biostratigraphic limits with the distribution of time-surfaces. These time-surfaces are established from sedimentological analysis and sequences stratigraphy. The Upper Cretaceous of the Gavrovo platform is formed by stacked shallowing-upward parasequences which are usually capped by an exposure surface. Most of them were exposed in supratidal environments and dolomitized to a greater or lesser extent. Some underwent continental diagenesis as recorded by karsts, microkarsts and karstic fillings (fig. 5), root traces, alveolar-septal structures, microcodiums, pseudomicrocodiums, pedogenetic pseudomicrokarsts and continental stromatolithic laminations. Although outcrops are great distances apart and located on different structural blocks, they record a major discontinuity within the Maastrichtian. It is characterized by continental exposure, a change in the main type of parasequences, and disruption of the parasequence stacking pattern as evidenced on Fischer plots. In each outcrop, limestones exhibiting continental diagenesis are cut by an erosional surface. This surface is proposed as a local maximum of regression and a transgressive surface. On a peculiar outcrop, the surface marks the inflection point between thinning-upward and thickening-upward parasequence trends on the Fischer plot. In proximal platforms, such a point can be interpreted as corresponding to a local maximum of regression and this surface is also a transgressive surface. The same features occurred in many other outcrops and show that the entire platform was subaerially exposed. Similar episodes of exposure associated with continental diagenesis are reported for Maastrichtian deposits of other Adriatico-Aegean platforms. Continental exposure and associated erosion is currently interpreted as a result of a fall in relative sea level caused either by uplift or by eustatic sea level fall. Successive shallowing up parasequences showing final exposure and continental diagenesis would imply an impossible yo-yo type subsidence. Accordingly, the proposed maximum of regression is thought to be eustatically controlled. Moreover, the maximum of regression caps CsB6 parasequences controlled by allocyclic mechanisms as indicated by similar stacking patterns in different and remote outcrops. This strongly suggests CsB6 sedimentation was eustatically controlled and the relevant maximum of regression is proposed as a time-surface. The distribution of foraminifer populations in the outcrops studied here indicates that the Rhapydionininae of biozone CsB6 do not occur above the maximum of regression. The regressive maximum clearly coincides with the disappearance of foraminifer species whereas the subsequent transgressive episode is characterized by the emergence of just one species. And yet, population renewal is not related to a fundamental change in the platform environment: very shallow water facies association below and above the maximum regression surface are identical. This supports the hypothesis that sea level variations were the cause of faunal extinction and renewal. It is evidenced too that the boundary between the two populations can be used as a time marker. In this case study, the biostratigraphy based on the use of benthic and shallow-water dependent foraminifera is genuinely chronostratigraphic.