An important task of plant population ecology is to understand what factors affect seedling recruitment, such as seed and microsite availability. Sowing experiments, i.e. seed addition into existing populations (augmentation experiments) or seed introduction into unoccupied habitats (introduction experiments), may provide unambiguous evidence for different types of limitation, such as seed or microsite availability limitation. Liaodong oak (Quercus wutaishanica), an endemic woody species, is a dominant species of the forest community in Mt. Ziwuling, but it has low natural regeneration rates and recruitment from acorns is difficult. In order to determine whether seedling recruitment in Q. wutaishanica is limited by seed availability, microsite availability, or both, a field sowing experiment was performed in 216 plots (30 cm×30 cm) in three typical stands (Liaodong oak forest, conifer plantation, and grassland, 72 plots per stand). The seed addition, disturbance (removing the litter) and shrub cover were selected as influencing factors. The numbers of emerged seedlings were tagged and monitored from April 2007 to October 2009. Light, soil moisture, and growth parameters (height, stem basal diameter and the wormholes per leaf) were measured bimonthly from April to October every year. Seedling data (number of oak seedlings per plot) in every stand were log-transformed to approximate normality. The effects of three treatments were analyzed with GLM Multivariate in SPSS version 13.0. The effects of treatments (seed addition and disturbance) on the number of recruited seedlings, variation of seedling growth across different years, and differences in light intensity and soil moisture among three stands were analyzed by One-Way ANOVA (Tukey HSD multiple comparison test). The results showed that, in the oak forest, leaf wormholes were significantly more numerous than in the conifer forest and the grassland. In the conifer forest, stem basal diameter was significantly higher than in the oak forest and the grassland. Furthermore, We found that seed addition and disturbance always significantly increased seedling emergence and recruitment in oak and conifer forests. There was an interaction between seed addition and disturbance. We demonstrated that recruitment of Q. wutaishanica seedlings was limited by seed and microsite availability in closed forests. Litter is one of the factors leading to microsite limitation in oak and conifer forests. However, in grassland, no treatment could improve oak recruitment, indicating that the open site was not suitable for seedling recruitment because of intense light and soil drought. It was also found that the survival rate of Liaodong oak seedling was significantly higher in forests than in grassland and that the conifer plantation was safer for seedling establishment according to the seedling growth in the three forest stands. Unexpectedly, shrub cover did not affect the recruitment of Liaodong oak seedlings. To conclude, we suggest that lack of viable seeds and suitable habitat are the main limiting factors for Q. wutaishanica recruitment in this region, and conifer plantation is the most suitable habitat for oak seedling survival. In early successional grasslands, Liaodong oak is unlikely to recruit successfully regardless of manipulations. We suggest that co-limitation by seed and microsite availability underlies the mechanism of Q. wutaishanica population recruitment.