Asian Soil Research Journal

Biofertilisers are promoted as low-input tools for increasing legume productivity, yet the category encompasses biologically distinct interventions whose field performance is not equally reliable. This critical narrative review evaluates whether microbial inoculants increase the productivity of grain and forage legumes, with particular emphasis on harvestable yield rather than nodulation or vegetative growth alone. Literature published principally from 1990 to 21 June 2026 was appraised alongside selected foundational studies. Evidence was synthesised across rhizobial inoculation, plant growth-promoting rhizobacteria, phosphorus-solubilising bacteria, arbuscular mycorrhizal fungi, Trichoderma and multi-organism co-inoculation. The strongest evidence supports compatible, effective rhizobial inoculation when indigenous rhizobia are absent, sparse, poorly competitive or inefficient and when other constraints to biological nitrogen fixation are not dominant. Large on-farm datasets and field experiments show meaningful average yield gains in soybean and chickpea, but also wide site-to-site variation. By contrast, non-rhizobial co-inoculants often increase nodulation, root growth or biomass more consistently than grain yield. Meta-analyses of soybean co-inoculation illustrate this distinction: improvements in intermediate traits may occur without a statistically reliable increase in harvested yield, while other datasets show only modest additional yield beyond conventional rhizobial inoculation. Fungal partners can enhance phosphorus acquisition, symbiotic functioning and disease suppression, but responses depend strongly on soil phosphorus, native microbial communities, crop genotype, water status and formulation quality. Across inoculant classes, controlled-environment effects tend to exceed field effects, indicating that colonisation, competition, environmental stress and agronomic co-limitations constrain translation. Biofertilisers therefore do increase legume productivity under identifiable conditions, but they are not universally effective substitutes for balanced soil fertility management. Future progress requires multi-site, multi-season trials with robust comparators, molecular tracking of inoculant occupancy, strain-by-genotype-by-environment matching, product-quality verification and endpoints centred on yield stability, nutrient-use efficiency and farm-level value.