Abstract: In this study, the sample data was based on 114 sample trees from dahurian larch plantations located in Qitaihe Forest Bureau in Heilongjiang Province, northeastern China. The Max and Burkhart (1976) segmented taper equation was fitted using nonlinear mixed-effects modeling techniques. The results showed that Max and Burkhart's model with random parameters b1, b2, b3 showed the best performance when the model was fitted using PROC NLMIXED module in SAS. Random parameter b1 showed negative correlation with crown ratio and crown length, and positive correlation with crown height. Random parameter b2 showed positive correlation with crown ratio and crown length, and negative correlation with crown height. Random parameter b3 did not show any correlation with crown ratio, crown length and crown height. Three taper equations including crown ratio, crown length and crown height variables were developed based on these correlations. Significant test of parameter estimates indicated that crown ratio, crown length and crown height had significant effects on stem taper. A series of taper curves were generated based on the stem taper equation with crown ratio. It shows that trees of the same relative size but with a larger crown ratio generally have a bigger stem taper, lower stem profile quality.