Tumor necrosis factor-a (TNF-a) is central to antimycobacterial immunity yet dysregulated signaling can promote immunopathology. Two TNF promoter variants, -308G/A (rs1800629) and-238G/A (rs361525), have been widely investigated for associations with tuberculosis susceptibility. This narrative review synthesizes evidence from 2000 to 2025 across diverse populations, integrating meta-analyses, case-control cohorts, and functional studies. Pooled analyses show no consistent overall association for either variant, with small and context-dependent signals in some subgroups. Functional assays indicate that -308A can increase promoter activity under defined stimuli, but genotype-phenotype relationships in patients are inconsistent and likely depend on cellular state, infection stage, and environment. Population differences in allele frequencies and linkage disequilibrium, together with variable study design and phenotyping, contribute to heterogeneous findings. Current data do not support clinical use of these variants as stand-alone biomarkers. The most promising applications involve research stratification within multi-ancestry, multi-omics frameworks and continued emphasis on latent tuberculosis infection screening before anti-TNF-a therapy. Priorities include adequately powered cohorts in high-burden regions, standardized phenotypes, and stimulus-aware functional genomics to resolve context-specific regulatory effects and guide pathway-focused interventions.