Bioengineering structures devoted to natural hazard mitigation often require the implementation of vegetation and inert materials, mainly timber logs. The latter are intended to ensure the protective function of the structure until the crop development of the plants is sufficient. The structural use of raw wood, i.e. without wood preservative treatment, in severe decay conditions is specific to the bioengineering structures. It raises practical and scientific issues regarding wood-decay assessment for structure management as well as for studies related to design improvements. In this study, two methods assessing the level of decay – the near infra-red spectroscopy (NIRS) method and the vibration resonance method (VRM) – developed on logs degraded in semi-controlled conditions were applied and compared. These methods are based on a mechanical definition of the level of decay by means of the normalized loss in the modulus of elasticity between intact and decayed states. The level of decay of 60 small-diameter silver fir logs degraded in natural conditions in the French Northern Alps was assessed using both methods. The samples had been put on the ground in four different sites near alpine torrents. The degradation process was assumed to take place in aerobic conditions. Prior to assessment of the level of decay, the applicability of the methods was verified. The level of decay was based on a classification varying from class 1 for the least decayed logs to class 4. Both methods allowed assessing the effects of the microbial activity on the logs. The levels of decay found were mainly classes 1 and 2 in both cases but to various extents depending on the method. Finally, the NIRS method was found to be less reliable than VRM. VRM can be used to assess the level of decay of logs decayed in natural conditions.