Abstract Rubberwood derived from plantation replanting of Hevea brasiliensis Müll. Arg. is an increasingly important renewable resource for timber, furniture, and biomass-based industries in Southeast Asia. Despite its growing industrial value, the genetic architecture underlying rubberwood quality remains largely unresolved, limiting the use of molecular breeding for simultaneous improvement of latex and wood traits. Here, we combined phenotypic evaluation, quantitative genetic analysis, and genome-wide association study (GWAS) to dissect the genetic basis of wood physical, mechanical, chemical, and energy-related traits in rubber tree. A panel of 100 rubber tree genotypes, including commercial clones and controlled-cross progenies, was evaluated using a 10 × 10 triple-lattice design with 300 sampled trees. Wood density, moisture content, modulus of rupture, modulus of elasticity, compressive strength, lignocellulosic composition, proximate traits, and elemental composition were measured and analyzed together with 24,242 high-quality DArTSeq SNP markers. Substantial phenotypic variation was detected for most traits, particularly for mechanical properties and cell-wall components, with broad-sense heritability ranging from 0 to 0.51. Moderate heritability for hemicellulose, hydrogen content, and compression parallel to grain indicated promising targets for genetic improvement and association mapping. Correlation analysis revealed biologically meaningful relationships among wood density, strength-related traits, and lignocellulosic composition, while population structure analysis identified two main genetic subpopulations. GWAS using five complementary models (GLM, MLM, CMLM, FarmCPU, and BLINK) identified 33 SNP-trait associations across 11 chromosomes, comprising 11 genome-wide significant associations exceeding the Bonferroni-corrected threshold and 22 suggestive associations. Several loci co-localized with genes involved in secondary cell wall biosynthesis, transcriptional regulation, and cellular metabolism. These findings provide the first integrated genomic framework for rubberwood quality in H. brasiliensis and establish molecular targets for genomic breeding of dual-purpose rubber clones with improved latex productivity and wood value.