Abstract: We established a 4 hm²(200 m×200 m) temperate secondary forest plot in the montane region of eastern Liaoning Province, China, and analyzed the spatial patterns and interspecific spatial associations of tree seedlings at scales of 0-50 m. The results showed that under complete spatial randomness null model, 95% of the species at the scale of 0-20 m showed a significant aggregation pattern, and 19 species at the scale of 0-16 m showed an aggregation pattern. The proportion of species exhibiting aggregated patterns decreased with the spatial scale. At the scale of 50 m, the random distribution pattern became the main form of tree species distribution. Underheterogeneous Poisson process null model, 5% of the species at the scale of 0-24 m showed a significant aggregation pattern, and 42% and 58% of the species at the scale of 26-50 m showed a random distribution and regularity patterns, respectively. Under complete spatial randomness null model, the proportion of species pairs of positive correlation was the highest, and at the scale of 50 m, the proportion of species pairs showed positive correlation, uncorrelated and negative correlation tended to be equal. Under heterogeneous Poisson process null model, species pairs mainly showed negative correlation, and the proportion of species pairs exhibiting negative correlation increased with spatial scale. The results implied that seed dispersal limitation and habitat heterogeneity could explain the aggregation pattern of tree seedlings. The strong aggregation of tree seedlings also contributed to close interspecific spatial associations and poor stability of the regeneration community.