Abstract： Larch(Larix olgensis) was one of the mostly planted tree species in East-northeast China. In this paper, spatial distribution pattern of larch natural regeneration seedling in 55a larch plantation dominated by Larix olgensis was analyzed.The coordinate, basal diameter,and height of natural larch seedling in plot (30 m×l8 m) were surveyed. Twenty-three kinds of contiguous grid quadrat sampling based on seedling spatial mapped points were simulated, and traditionally data-mining methods of spatial pattern were utilized to find spatial patterns at twenty-three kinds of discrete scale. Traditionally data-mining methods of spatial pattern in this paper included: utilizing discrete poisson, negative binomial and binomial distribution functions to fit seedling discrete distribution; variance to mean ratio, Morisita index and nearest neighbor analysis were used to confirm seedling pattern type; utilizing C_A index, Green index and K value of negative binomial distribution to quantify pattern intensity. Semivariogram analysis and spatial point pattern analysis (Ripley's second order analysis method) based on seedling spatial mapped points were utilized to find the change of spatial pattern type and pattern intensity at continuous scale, and patch size. Seedling spatial arrangement by seedling nearest neighbor distance from seedling and from old-tree was studied to find larch natural regeneration inhabitancy character. Seedlings were clumped at all discrete scales (quadrat area, 1-18 m~2). At 0-10.5m (radius range) continuous scale, seedling pattern was aggregate pattern, but at 10.5-12.3 m (radius range) continuous scale, seedling pattern was random pattern. At all discrete scales (quadrat area froml m~2 to 18 m~2), aggregate intensity indexes increased with scale increasing. At 0-10.5 m (radius range) continuous scale, aggregate intensity first increased with scale increasing, and then declined at greater scales. At 4.2 m (radius) scale, aggregate intensity was the maximum. The biggest radius of patch, which was formed by natural regeneration seedling, was 4.79 m. The seedling nearest neighbor average distance between seedling and rom old-tree were 030-0.42 m and 2.99-3.26 m, respectively. The results from contiguous grid quadrat sampling method and spatial point pattern analysis method were very similar on the whole in their abilities to discriminate spatial pattern and quantify aggregate intensity, but spatial point pattern analysis method was superior in its ability to detect changes of spatial pattern and aggregate intensity with scale continuously chang-ing. The combined utilization of spatial pattern methods based on spatial mapped points, such as spatial point pattern analysis, nearest neighbor analysis and semivariogram analysis, would provide abundant information about population pattern.