Based on biomass measurements of 11 major tree species at five stand age classes (young, middle-aged, near-mature, mature, and over-mature forests), the optimized allometric biomass of various tree species was modeled. The total biomass of forest ecosystems in Guangxi, as well as the biomass composition and contribution in various forest types for various stand ages and forest layers were investigated. The leaf, branch, stem, root, and total single-tree biomass for 11 tree species, except Cunninghamia lanceolata and Eucalyptus urophylla × grandis leaf to diameter at breast height (DBH) (D), the ratio of belowground to aboveground biomass, and the tree height to DBH ratio (D) were best fitted with a power regression model at a significance level of P < 0.05 using t-tests. The best fit was observed for the total single-tree biomass to DBH (D) for the 11 tree species. The total forest biomass in Guangxi was 1425.7 Tg, and the average forest biomass was 105.36 Mg/hm². The total stand biomass of major tree species was ranked in the following order: Pinus forest (366.14 Tg) > hardwood forest (291.08 Tg) > softwood forest (239.75 Tg) > karst forest (165.51 Tg) > Cunninghamia lanceolata forest (164.01 Tg) > Eucalyptus forest (99.55 Tg) > Quercus forest (46.34 Tg) > Octagon forest (20.21 Tg) > oil-tea Camellia forest (19.59 Tg) > Bamboo forest (13.19 Tg); the biomass of each forest increased with stand age. The biomass of the overstory tree layer accounted for 78.30% to 97.47% of the total forest biomass, indicating that the overstory tree layer dominated the total biomass. Moreover, the aboveground biomass was greater than the belowground biomass at various forest ages. Considering over-fitting of the statistical models, the explainable proportion of variance, and the significance of regression coefficients in the allometric models, it was demonstrated that the mathematical model of biomass, using DBH (D) as the single variable, could effectively predict biomass for the main forest species as well as the total forest biomass in Guangxi. The optimized allometric models and the stimulate values of the ratio of aboveground to belowground biomass were of great value for estimating the belowground biomass of primary wood species in Guangxi.