1) Introduction In the static and dynamic workflow of carbonate reservoirs, stratigraphic correlation of well data is one of the first and most influent steps. Indeed, facies distribution and petrophysical properties mainly control flow simulation and are often computed thanks to geostatistical methods, on grids based on stratigraphic correlation and structural data interpreted from seismic data (Borgomano [2008]). In reservoir uncertainty modeling approaches, a unique grid is built, and uncertainties about layering geometry, facies distribution and petrophysical properties are handled using multiple geostatistical simulations (Charles et al, 2001). This article aims at assessing uncertainties due to stratigraphic correlations by also generating several set of possible stratigraphic well correlations. Several grids may then be built from these results and used for facies and property modeling. The method presented here generates automatically and stochastically sequence stratigraphic correlations of carbonate ramp systems by hierarchically integrating multiple pieces of 3D information as: interpreted well data,correlation lines extracted from seismic, andinformation obtained on analogs. 2) 3D Stochastic Stratigraphic Well Correlation method To perform the correlation, we propose a multi-dimensional and stochastic extension of the Dynamic Time Warping Algorithm (DTW, Myers et al., 1981) that we call msDTW. The DTW algorithm provides a way to find the optimal alignment between two time series [Myer et al., 1981]. This algorithm was used for the correlation of two wells by Smith and Waterman [1980], Howell [1983], Waterman and Raymond [1987], Griffiths and Bake [1990], Brown [1997] for example. Lallier et al. [2009] presented an improvement of the DTW, making the method stochastic and introducing a hierarchy to mimic the reasoning made by sedimentologists when correlating well data (Fig. 1), and applied this method to a carbonate ramp system.