Mediterranean dry grasslands are characterized by a high biodiversity which is threatened by significant changes in land-use in the last decades. The La Crau area is the only French Mediterranean steppe located in South-Eastern France. From the 45,000ha occurring in the 16th century, only 11,500ha are currently remaining. Moreover numerous studies have shown a very low resilience of this plant community after major disturbances, such as cultivation, even on the very long term (150 yrs). This has stressed both the importance of conserving remnant patches and restoring plant community. Several experiments testing species-transfer methods have been carried out in the last decade. In 2002 the transplantation of two perennial target species: Thymus vulgaris and Brachypodium retusum in former cultivated fields showed that their establishment was successful if the stone cover had been previously restored. The transfer of vacuum-harvested hay was assessed in 2007 on an area disturbed by pipeline burying. This method allowed the plant composition to be close to that of the reference community, with many target annual species, but without some dominant perennial species. In 2009, a multi-treatment experiment was implemented on a larger area disturbed by intense peach cultivation where hay transfer and soil transfer, combined or not with topsoil removal were experimented. The best results were obtained when both abiotic conditions were restored (topsoil removal, soil transfer) and species were transferred (soil transfer, hay transfer combined with topsoil removal). In 2011, large scale soil transfers with or without soil compaction were implemented again on an area degraded by an oil spill where soil had been excavated. On the very short term, soil transfers showed a species-rich community composed of numerous target species. Their combination with soil compaction or topsoil removal allowed decreasing the sometimes too dense cover of grasses occurring on other restoration treatments. Results from these ten years of diverse experiments and monitoring highlight the difficulties of restoring this unique habitat. Nevertheless they provide insights on possibilities of approaching the target communities, especially when the transfer of the whole community is combined with the restoration of appropriate abiotic conditions.