where 13R stands for the ratio 13C/12C, standard being PDB (marine carbonate: belemnite from the Pee Dee Formation). Isotopic fractionation of carbon associated with photosynthesis in living terrestrial plants depends upon three main factors (e.g. O'Leary, 1981): (1) the photosynthetic pathway of the plant, C3, C4 or CAM, (2) the inorganic-carbon source and (3) environmental parameters. Such factors can be recorded by fossil plant via 613C values. Carbon isotope ratios were often used to investigate the emergence of C4 photosynthetic pathway in fossil plants. Stable carbon isotopic abundances were also studied to research an isotopic record of palaeoatmospheric condition in terrestrial fossil plants, to reconstruct local palaeoenvironments or as chemostratigraphic tool. However, stable carbon isotope abundances in fossil plants can be diagenetically altered if the carbon lost by degradation has 613C values significantly different of the carbon which is preserved. Hedges et al. (1985) showed that lignin is selectively preserved to the detriment of cellulose in buried wood. Since lignin is 13C-depleted when compared to the whole tissue, that may lead to significant changes in 813C values as demonstrated by Benner et al. in a modern salt marsh (1987). However, several experimental studies on modern plants (e.g. Balesdent and Mariotti 1996) and investigations on fossil ones suggested that fossil plants typically have retained their original isotopic composition. That could be explained by the presence of remnant of carbohydrates in fossil plants. Nevertheless, significance of ~13C values in fossil plants is strongly dependent on the level of diagenesis. The aim of the present study was thus to get a tool allowing for investigation of the isotopic signal in fossil plant whatever the level of preservation of bulk ~13C values. Isotopic analysis of an individual molecular biomarker, present in substantial amounts in fossil plants and stable through geological time, could constitute an adequate methodology. The identification of such a biomarker was first undertaken in leaf lipids since most lipid constituents are stable through geological times, and, were shown not to move into the surrounding sediment during burial (Logan et al., 1995). To validate the use of a given biomarker it was necessary to measure its 513C value in fossil plants in which the bulk isotopic signal is well preserved, and, to compare these values with those obtained in modern counterparts. The Cenomanian lagoonal Member of the 'Argiles du Baugeois' (Anjou, France) yielded a rich and exceptionally well preserved fossil flora. Preliminary studies have shown that the bulk ~13C values were not significantly altered in these fossil plants. 13C/12C ratios thus allowed the reconstruction of the ecological distribution of each species studied, along a decreasing salinity gradient from the lagoonal zones to the flood plain. Leaf lipids from the most abundant species of the deposit, the Ginkgoale Eretmophyllum andegavense, were thus analysed, and compared to the ones of its extant counterpart Ginkgo biloba. GC-MS analyses of leaf lipids from both Ginkgoales reveal the presence of 3 common compounds in E. andegavense and its modern counterpart: long chain n-alkanes, predominantly even n-acids and alkyl, dimethoxycoumarins. They all constitute potential biomarkers of the isotopic signal. Isotopic analyses were first undertaken on n-