The liana species Derris trifoliata is an associate species in mangroves, but in recent years, its biomass and abundance have increased. The biomass estimated for Derris trifoliata is vital for determining the carbon stores in mangrove habitats. And the allometric equations are insufficient but essential for lianas. We collected 132 samples of D. trifoliata to produce a regression analysis, following which allometric equations of biomass (root, stem, leaf, ground, and total biomass) were obtained. The allometric equations we obtained based on basal diameter have a better effect to predict stem biomass, ground biomass, and total biomass, and the adjusted R-2 for these measures was 0.5258, 0.5287, and 0.5397, respectively. When leaf biomass was simultaneously added as another independent variable, the accuracy of the prediction of the total biomass (adjusted R-2 = 0.6715) and stem biomass (adjusted R-2 = 0.6163) improved. The ratio of root biomass to total biomass, stem biomass to total biomass, and leaf biomass to total biomass was 0.208-0.430, 0.394 to 0.643, and 0.146 to 0.199, respectively. The biomass accumulation increased with the increase of basal diameter and the ratio of root biomass to ground biomass decreased with the increase of basal diameter, which is likely due to the limited light resources. We measured the stem diameter where the raceme was located and a total of 304 raceme or fruit spur samples were selected. Finally, we found that the number of raceme or fruit spur increased first and then decreased with the increase of stem diameter. The stem diameter, to some extent, represents the age class of the stem: thinner stems have a greater chance of sexual reproduction than thicker stems. The models provided in this paper will facilitate the estimation of the productivity of forests and the amount of carbon sequestered by forests, and the relationship between the generation of inflorescences and stem diameter can predict sexual reproduction and diffusion in the present D. trifoliata distribution area. (C) 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).