The evolution of atmospheric CO2 and the pH of the ocean can be reconstructed by the boron isotopic composition (δ11B) of marine carbonates, which is influenced by the δ11B of the seawater. Boron (B) in the ocean is primarily affected by continental weathering through rivers. Thus, it is essential to understand the behavior of B and B isotopes in rivers and the factors affecting riverine B, which require a better understanding of sources and processes of B in river systems. This review evaluates the inventories of B reservoirs contributing to rivers and investigates the processes regulating the B isotope geochemistry of rivers. B is widespread at the Earth’s surface and shows a wide range of concentrations between reservoirs. Different reservoirs also exhibit significant variations in B isotopic compositions. Mixing and Rayleigh effects are mainly responsible for the variations in the δ11B values of meteoric precipitation, which result in marine (δ11B = +37 ± 7‰), anthropogenic (δ11B = +9 ± 10‰), and mixing types (δ11B = +17 ± 13‰) of meteoric precipitation. The contribution of B to rivers from carbonate dissolution is negligible. Marine and non-marine evaporites have distinct δ11B values (marine δ11B: + 27 ± 9.4‰ and non-marine δ11B: −2 ± 8.6‰) that primarily reflect their different depositional environments. S-type granites that are tourmaline-free have an estimated δ11B value of −14.2 ± 4.9‰ and a Na/B value of 140 ± 34. Non-S-type granites have a δ11B value of −8.9 ± 6.7‰ and a Na/B value of 1190 ± 170. Intraplate basalts exhibit a δ11B value of −5.2 ± 4.4‰ and a Na/B value of 3300 ± 770. Subduction-related basalts have a δ11B value of + 0.3 ± 7.3‰ and a Na/B value of 1060 ± 830. Shale has high B contents of siliciclastic sedimentary rocks (104 ± 92 ppm). The inferred δ11B values of marine and continental shales are −8‰ and −16‰, respectively. The effects of metamorphism can vary widely depending on the geologic setting and type of protolith. The δ11B values of wastewater are investigated based on their industrial, agricultural, and urban sources. This inventory of B reservoirs can be useful for studies on rivers on a continental scale. In regolith and groundwater, B isotopic fractionation mainly occurs due to water-rock interactions and the biological cycle of B, whereas adsorption on sediments leads to minor B isotopic fractionation in rivers. In groundwater, the reactive transport model reveals that the δ11B value of river water is sensitive to hydrological conditions. In regolith, the steady-state mass balance model is used to predict the B isotope behavior of soil solution in different weathering regimes. In the supply-limited regime (where chemical weathering is limited by tectonic forcing), the precipitation of secondary minerals controls the variations in the δ11B values of soil solution, leading to an increase in the difference in the δ11B values between soil solution and parent rock (δ11Bdiss−δ11Brock) with lower denudation rates, whereas secondary mineral dissolution produces the opposite change in δ11B. In the kinetically limited regime (where chemical weathering is limited by climate), the biological cycle controls the variations in the δ11B values of soil solution, and the δ11B values of soil solution generally become closer to those of parent rock with higher denudation rates. The relationship between the denudation rates and δ11Bdiss−δ11Brock is thus not monotonous, indicating that additional constraints are required to distinguish between the two regimes. Understanding of B isotope geochemistry of rivers can be improved by better constraints on B end-member estimates, investigation of the B isotopic fractionation caused by weathering and biological cycling in regolith, and assessment of atmospheric and biological sub-cycle.