Ecological interactions of species and thus their spatial patterns may differ between homogeneous and heterogeneous forests. To account for this, techniques of point pattern analysis were implemented on mapped locations of tree individuals from two 1-ha tropicalforest plots in Vietnam. We analyzed the effect of environmental heterogeneity on tree distributions; spatial distribution patterns of dominant species; inter-specific associations; and conspecific associations between life stages. Our analyses showed that: environmental conditions were homogeneous at plot 1 but heterogeneous at plot 2; in both plots, all six dominant species were aggregated at various scales up to 30 m, and tree species were aggregated at larger scales in the homogeneous site than in the heterogeneous site; attraction between pairs of species was remarkably higher at the homogeneous site while negative associations were more frequent in the heterogeneous site; some species, H. kurzii, T. ilicifolia(homogeneous plot) and D. sylvatica, S. wightianum(heterogeneous plot),showed a lack of early life-stage individuals near conspecific adults. Moreover, additional clustering of young individuals was independent from conspecific adults, except D. sylvatica in both sites. These findings are consistent with the Janzen-Connell hypothesis. Overall, habitat heterogeneity influences spatial patterns and inter-specific associations of the tree species and evidences of self-thinning are shown in most species. Ecological interactions of species and thus their spatial patterns may differ between homogeneous and heterogeneous forests. To account for this, techniques of point pattern analysis were implemented on mapped locations of tree individuals from two 1-ha tropicalforest plots in Vietnam. environmental heterogeneity on tree distributions; spatial distribution patterns of dominant species; inter-specific associations; and conspecific associations between life stages. Our analyzes are that: environmental conditions were homogeneous at plot 1 but heterogeneous at plot 2; in both plots, all six dominant species were aggregated at various scales up to 30 m, and tree species were aggregated at larger scales in the homogeneous site than in the heterogeneous site; attraction between pairs of species was remarkably higher at the homogeneous site while negative associations were more frequent in the heterogeneous site; some species, H. kurzii, T. ilicifolia (homogeneous plot) and D. sylvatica, S. wighttianum (heterogeneous plot), showed a lack of early life-stage individuals near conspecific adults. Moreover, additional clustering of young individuals was independent from conspecific adults, except D. sylvatica in both sites. These findings are consistent with the Janzen-Connell hypothesis. Overall, habitat heterogeneity influences spatial patterns and inter-specific associations of the tree species and evidences of self-thinning are shown in the most species.