This thesis has been submitted to the Department of Geoscience at the University of Oslo in accordance with requirements for dissertation for the degree of Philosophiae Doctor. The work that forms the basis for this Ph.D. thesis was conducted under the supervision of Prof. Dr. Dag Arild Karlsen and financed by NORECO ASA. Although first drillings were conducted during the early 1980ies, the Barents Sea is still considered a frontier area for petroleum exploration. Despite extensive research on the geologic evolution, sedimentology and stratigraphy, geochemical studies reporting on petroleum systems are scarce. It has been suggested relatively early in the exploration history that oils and condensates in the region could be regarded as being mixtures of more than one filling event. Variable results in the early part of the exploration phase are undoubtedly related to the complexities of the region. It is commonly accepted that several phases of uplift, erosion and glacial events during the Cenozoic had great impact on petroleum system elements, but also on already accumulated petroleum. Generation from miscellaneous source rocks, changes in pressure-volume-temperature conditions, secondary inreservoir alteration, large scale remigration and entrapment mechanisms, and leakage of petroleum are among the processes that impede petroleum system investigation. The purpose of this work was to provide systematic evaluations of the degree to which oils and condensates are "blends", or of singular source rock origin, and to evaluate potential variations in maturity signatures, biodegradation, migration induced phase-fractionation and source rock facies. Realizing that petroleum geochemical studies in the Barents Sea may be complicated due to extensive alteration and blends of oils in traps, an attempt was made to decipher the complex signatures: A full geochemical fingerprint of each sample in terms of thermal maturity, secondary alteration effects, age, paleo depositional environments and organic matter input had to be created. Therefore, systematic analysis of three hydrocarbon compound classes has been applied: (1) light hydrocarbon C 4-C 8 compounds, (2) medium range C 10-C 20 compounds, and (3) biomarker range C 20+ compounds. The results indicate petroleum generation from the early oil window to the late oil/ condensate window. Phase fractionated condensates and oils have been observed in the western part of the Hammerfest Basin. Petroleum mixtures have been identified by varying thermal maturities among the three different compound classes, and paleo biodegradation signatures in combination with fresh, unaltered charges. This indicates at least two migration events of highly variable maturity and/or even source rock facies signatures. Similar geochemical characteristics and use of multivariate statistical analysis led to classification of four petroleum families: (1) Family A: Permian/Triassic sourced, (2) Family B: Carboniferous sourced, (3) Family C: Jurassic sourced, and (4) Family D: Triassic and Jurassic sourced...