The processes involved in the formation of petroleum have been debated for approximately 100 years and numerous opposing views still exist. Some believe that petroleum forms directly from the source rock kerogen, while others believe that it forms from heavy asphaltic products (NSOs) generated early in the maturation process. Much of the uncertainty lies in the fact that, until now, there have been few methods capable of obtaining molecular-level information on generally insoluble kerogen and associated asphaltic NSO materials using non-invasive methods. Our work brings some new non-invasive NMR and mass spectrometric techniques that bear on testing the possible pathways for petroleum formation. Our overarching hypothesis is that we can resolve the issue regarding the major pathway of petroleum formation by a comparison between the chemical structure of NSOÕs generated from simulations of petroleum formation and the chemical structure of their precursor kerogen at a high molecular-scale specificity that has not been attempted before. The NMR methods allow us to establish that soluble extracts are representative of the whole kerogen. The chemical and structural comparison between the initial kerogen and solvent extracts of this kerogen has been evaluated using 2D HSQC NMR techniques. This approach allows a direct comparison of insoluble residue with soluble extracts. The insoluble residue was examined by HRMAS NMR and the extracted materials were examined by traditional liquids NMR. Two extraction approaches have been used for a type I kerogen from the Mahogany Zone of the Green River Formation. The first approach involved three successive extractions using three different solvents (n-pentane, DCM and pyridine) and the second employed a single solvent extraction using pyridine. The result of this work has been published in Organic Geochemistry (Salmon et al., vol.42, p. 301-315). Using quantitative solid state 13C NMR and 2D HSQC NMR techniques, we showed that the combined chemical composition of the three successive extracts was similar to the composition of the kerogen and of a single pyridine extract. These findings confirmed that a representative aliquot of the kerogen can be extracted with either approaches. The extracts were analyzed by Fourier transform mass spectrometry, a technique that is so precise in mass assignment that unique molecular formulas can be calculated. We observed that the peaks of some of the main series, detected in the successive extracts analyzed by FT-ICR-MS, were only weakly observed in the single pyridine extract. These observations are probably the results of the charge competition effect, not encountered in the successive extracts. Therefore, for the present work, we use the successive extraction approach as it appears to be most representative of the organic matter. In order to examine the molecular and structural relationships between the kerogen and the products, we conducted low-level maturation studies in the laboratory of F. Behar at IFP (Institut Francais du Petrole) in France by closed-system pyrolysis experiments that simulate long-term maturation of kerogen. We are currently using the analytical strategies described above to establish relationships between the initial kerogen, its solvent-extractable products, and the residual kerogen. Artificial maturation experiments that involve two other kerogen (a type II and a type III) are conducted at IFP. The type I kerogen, after mineral disruption, was extracted successively with n-pentane, and DCM. Several series of polar CHO, CHOS and CHON compounds between C12 and C50+ were found in the initial kerogen analyzed by FT-ICR-MS. In both extracts, only 15 homologous series that represented at least 72% of the total peak area of each spectrum were identified. In the two types (n-pentane and DCM extractable) of NSOs generated during the artificial maturation and analyzed by FTICR-MS, similar series of polar CHO, CHOS and CHON compounds were found; however their molecular weights tend to decrease as a function of increasing severity of thermal maturation calculated as % conversion. Compounds containing only CHO elements in their make-up were dominant in each spectrum and were comprised mostly of the series CnH2nO2, CnH2n-2O2, CnH2n-2O4, and CnH2n-10O2 which all correspond to carboxylic acid-containing compounds. One chemical pathway that has been proposed for the generation of hydrocarbons found in petroleum is decarboxylation of acid compounds. For the first time here we have directly analyzed by negative ionization FTICR-MS and have characterized the distribution of carboxylic acid compounds extracted from the kerogen without any alteration or derivatization. We show that the distribution of carboxylic acid compounds obtained by ESI-FTICR-MS may be directly correlated to the distribution of hydrocarbons using the carbon preference index (CPI). By calibrating the FTICR-MS with internal standard, it was possible to assign a response factor for the carboxylic acids and dicarboxylic acids, and consequently get better quantification of the total mono- and di-carboxylic acids contents. This response factor is invariant in the carbon number range of C20 to C40. Also, after calibration, the carboxylic acid yield remains dominant among the other peaks detected by FTICR-MS. The quantitative proportions of these carboxylic acids compounds in the NSOs are then compared to the proportion of hydrocarbon produced. We have demonstrated in this work this first year that the molecular-level structural relationship between kerogen and the associated NSO compounds naturally formed and produced by artificial maturation under laboratory conditions. In our second year we will focus on the two other types of kerogen which are recognized as representative of the major petroleum systems on Earth. The differences between these three types of kerogen have a direct impact on both the quantity and the quality of petroleum. It is crucial to accurately describe the chemical structure of the kerogen and the thermally produced NSO products as well as the volatile hydrocarbons in artificial maturation experiments to determine the main reactions occurring during natural decomposition under geothermal stress and to establish the role of NSO compounds in this process. Our observations for type I kerogen that a genetic link between fatty acids and hydrocarbons exists: these lead us to consider that a simple decarboxylation reaction may be the most important thermal degradation reaction in petroleum formation.