Seep carbonates often precipitate at the sea floor or in the very shallow subsurface due to bio-chemical oxidation of hydrocarbons, rising from depth, coupled with the reduction of sulfates contained in sea water. Seep carbonates may thus be regarded as (fossil) indicators of hydrocarbon leakage, and are often used as such in the geological record where they are the most obvious remains of this process. Based on this paradigm, the aim of this project is to explore the possibilities to use seep carbonates to reconstruct the activities of a petroleum systems, i.e. the concept unifying the essential elements of petroleum geology (source rock, reservoir, seal, overburden) and processes (trap formation, generation, migration, accumulation). The ultimate objective of this research, in collaboration with TOTAL SA, is to integrate the identification of seep carbonate on seismic records as a tool for hydrocarbon exploration. This work is mostly based on the study of outcrop analogues in California, Taiwan, and France in combination with classic techniques of field geology, petrography, mineral and organic geochemistry, mapping, and basin modeling.