For elucidating effective methods of restoration of forest recreational sites where management goals are maintaining naturalness and conserving natural ecosystem, seeding, fertilization and soil surface treatment were used for four years at the devastated forest-floor. And seeding treatment was used for 2 years at an artificial bare ground for selecting tree species suitable for forest-floor revegetation. For restoration of forest-floor vegetation, factorial experiment was used with a split plot design (main plot : fertilization, subplot : soil surface×seeding) and a randomized complete block design(fertilization × seeding) at the Kwanaksan Aboretum, Anyang, Kyonggido. For selection of tree species suitable for forest-floor revegetation, shading, soil surface softening and straw-mat mulching treatment were used with a randomized complete block design using 19 native tree species in the same factorial design . Results were summarized as follows 1. Soil surface softening with tipping and ripping and straw-mat mulching(70% coverage) treatment was effective on germination, survival and growth of seeded vegetation at devastated forest -floor. Expecially, straw-mat mulching treatment was effective on soil surface stabilization and seedling`s survival at eroded soil surface, while complete soil surface softening treatment was effective on germination, survival and early growth of tree species(Cornus controversa, Euonymus sieboldianus, etc. ) of late-successional seres. 2. Introducing seeds of native species(Zanthoxylum schnifolium, Styrax japonica, etc.) of pioneer or early-successional seres, with good growth capability in barren soil was effective on rapid restoration in devastated forest-floor with its soil surface previously compacted and its surviving seeds washed away. 3. When the seeding and straw-mat mulching after partial soil surface softening with tipping and ripping treatment were employed, it took about three years to restore the devastated forest-floor. Where surface erosion had been undertaken for an extended period of time and where naturally surviving seeds of native species had been washed away. 4. Softening treatment of soil surface was effective for about two years, and seeding and soil surface treatment increased number of seelings and improved soil surface environment through fixing of movement of the fallen leaves. 5. The effect of Lespedeza maximowiczii used as a nitrogen fixing plant was not observed due to severe surface erosion after seeding operations. Direct seeding and subsequent growth of Quercus acutissima were not noticed due to vandalism by widelife as well as man. 6. Fertilizing effect was not oberserved, mainly due to seedling exposure and poor physical condition including soil surface erosion, low soil water potential and drought, etc, at the field experimental site. However, application of nitrogen and phosphate fertilizers was effective on seedling survival of the species in late-successional seres, while lime application, adversely affected the seedling survival. 7. Native species suitable for restoration of bareland under 75% shading in central part of Korea are Styraz japonica, Styrax obassia. Smilax china, Callicarpa japonica, Stephtznaradra incise, Viburnum dilatatum for. pilosulm, Magnolia sieboldii, Cornus kousa, Celastrus orbiculata, etc. Especially, Magnolia sieboldii, Stephanandra incisa and Styrax obassia were tolerant for forest-floor with hardened soil surface and Callicarpa japonica, Viburnum dilatatum for. pilosulm, Euonymus sieboldianus and Philadelphus schrerkii suitable for softened soil surface with straw-mat Philadelphus schrenkii suitable for softened soil surface with straw-mat mulching. 8. The seeds of Styrax obassia, Styrax japonica, Viburnum dilatatum for. pilasulm, Smilax china and Stephartandra incisa needed to break seed dormancy to accelerate germination.