Batch sorption experiments at fixed initial Np(V) concentration (∼1 × 10−6 M 237Np), M/V ratio (4 g L−1), and ionic strength (0.1 molal NaNO3) were conducted to determine the effects of varying pH and $P_{C{O}_2}$ on Np(V) sorption on SAz-1 montmorillonite. The results show that Np(V) sorption on montmorillonite is strongly influenced by pH and $P_{C{O}_2}$. In the absence of CO2, Np(V) sorption increases over the entire pH range examined (∼3 to ∼10), with measured sorption coefficients (KD) of about 10 mL g−1 at pH < 6 to KD ∼ 1000 mL g−1 at a pH of 10.5. However, for experiments open to atmospheric CO2 ($P_{C{O}_2}={10}^{-3.5}atm$), Np(V) sorption peaks at KD ∼ 100 mL g−1 at pH of 8 to 8.5 and decreases at higher or lower pH. A comparison of the pH-dependence of Np(V) sorption with that of Np(V) aqueous speciation indicates a close correlation between Np(V) sorption and the stability field of the Np(V)-hydroxy complex NpO2OH0 (aq). In the presence of CO2 and aqueous carbonate, sorption is inhibited at pH > 8 due to formation of aqueous Np(V)-carbonate complexes. A relatively simple 2-site Diffuse-Layer Model (DLM) with a single Np(V) surface complexation reaction per site effectively simulates the complex sorption behavior observed in the Np(V)-H2O-CO2-montmorillonite system. The good agreement between measured and DLM-predicted sorption values suggests that surface complexation models based on parameters derived from a limited set of data could be useful in extrapolating radionuclide sorption over a range of geochemical conditions. Such an approach could be used to support transport modeling and could provide a better alternative to the current use of constant KD values in performance assessment transport calculations.