Several microcosm experiments were run in parallel to evaluate the efficiency of microbial mats for crude oil degradation as compared with physico-chemical weathering. The oils used in the experiments constituted representative examples of those currently used for commercial purposes. One was aliphatic and of low viscosity (33.4 American Petroleum Institute degrees, degrees API) and the other was predominantly aromatic, with high sulphur content (ca. 2.7%) and viscosity (16.6 degrees API). After crude oil introduction, the microcosms were kept under cyclic changes in water level to mimic coastal tidal movements. The transformations observed showed that water weathering leads to more effective and rapid elimination of low molecular weight hydrocarbons than microbial mat metabolism, e.g. n-alkanes with chain length shorter than n-pentadecane or n-heptadecane, regular isoprenoid hydrocarbons with chain length lower than C16 or C18 or lower molecular weight naphthalenes. Microbial mats preserved these hydrocarbons from volatilization and water washing. However, hydrocarbons of lower volatility such as the C24-C30 n-alkanes or containing nitrogen atoms, e.g. carbazoles, were eliminated in higher proportion by microbial mats than by water weathering. The strong differences in composition between the two oils used for the experiments were also reflected in significant differences between water weathering and microbial mat biodegradation. Higher oil viscosity seemed to hinder the former but not the later.