In Western Europe, recreational amenity is presented as an important cultural ecosystem service that, along with other values, helps justify policies to conserve biodiversity. However, whether recreational use by the public is enhanced at protected areas designated for nature conservation is unknown. This is the first study to model outdoor recreation at a national scale, examining habitat preferences with statutory designation (Site of Special Scientific Interest) as an indicator of nature conservation importance. Models were based on a massive, three year national household survey providing spatially-referenced recreational visits to the natural environment. Site characteristics including land cover were compared between these observed visit sites (n = 31,502) and randomly chosen control sites (n = 63,000). Recreationists preferred areas of coast, freshwater, broadleaved woodland and higher densities of footpaths and avoided arable, coniferous woodland and lowland heath. Although conservation designation offers similar or greater public access than undesignated areas of the same habitat, statutory designation decreased the probability of visitation to coastal and freshwater sites and gave no effect for broadleaved woodland. Thus general recreational use by the public did not represent an important ecosystem service of protected high-nature-value areas, so that intrinsic and existence values remain as the primary justifications for conservation of high nature value areas. Management of 'green infrastructure' sites of lower conservation value that offer desirable habitats and enhanced provision of footpaths, could mitigate recreational impacts on nearby valuable conservation areas.