The spatial distribution and association of populations can reflect succession patterns and its adaptation strategies to the change of environmental factors, with important significance for vegetation restoration, reconstruction, and biodiversity conservation. We used the point-pattern analytical method to examine the spatial distribution and its association of the Loropetalum chinense population in karst hills of Guilin, China, based on field surveys. On the basis of Ripley K function, we used the pair-correlation function statistic derived to explore difference in the distribution patterns.We used Ripley L function to examine the spatial associations among the three diameter classes. The results showed that the diameter class structure of the population exhibited an irregular inverted "J" type, with the small diameter class occupying a large proportion, indicating an increased population structure with good regeneration capability. The individuals of the three diameter classes had a clumped distribution at the small scale. With the increases of spatial scales, the aggregation intensity gradually weakened and tended to be randomly distributed. There was an independent spatial association between individuals among the three diameter classes at small scale. As the scale increased, the spatial association between individuals with different diameter classes became positive or negative association. The greater the difference in diameter class of L. chinense population, the weaker their spatial correlation was, which might gradually turn into negative association. Our findings contributed to a clearer understanding of the ecological strategies and scale-dependent cha-racteristics of species coexistence and underlying mechanisms during the growth and development of L. chinense population in karst hills of Guilin, providing reference for the restoration, reconstruction, protection and management of forest in karst hills.