Planning applications for surface coal mine sites in South Wales are progressively being scrutinised more in respect of biodiversity mitigation and enhancement schemes for sustainable mining. In the mid-to late-1990s a commitment to establishing an often undefined type of semi-natural grassland was sufficient. This progressed to the specification of the target type of grassland in the next decade. Recent applications are seeing the planning authority and the statutory advisers demanding details of how this will be achieved and the evidence that this is possible. Given that ratcheting of the demands is a recent experience, it is difficult for the mining companies to provide the evidence on demand in the absence of a knowledge base. The availability of local seed is one of those details now being demanded. It is a particularly insidious topic if the information is not available (as it is the basis of any grassland re-establishment proposal), and without a satisfactory answer the proposal will fail and might prejudice the coal project as a whole. The applicant needs to have a primary dedicated resource and a backup resource available, and the knowledge and ability to exploit these resources. The State coal company, British Coal, in the 1980-90's commissioned a number of research programs. One of these was an in depth investigation in the seed harvesting of grassland and fen meadow vegetation in South Wales. The results of this unpublished investigation of the early 1990's are now germane. The objective of this paper is to set out the lessons learnt from this previous work for both the practitioner and regulator. One of the findings was more than one harvest in any growing season will be required for the full range of species, and this might span the period from June to October in some cases. This has implications for the harvesting strategy, whether it is a one year seeding or multi-year seeding program, how the seed resource is harvested spatially, and cognisance of the year to year variation in viable seed production due to weather and insect damage. The recommendations suggested in this Case Study may be beneficial elsewhere where harvesting local seed is necessary for successful reclamation.