Nitrate is a common groundwater pollutant all over the world. In some regions of Iran, its levels are high enough to cause serious problems to human health and the environment. The objectives of this work were to evaluate the efficiency of Fe-doped TiO2 nanoparticles at removing nitrate from aqueous solutions under UV and solar radiation and to model nitrate removal using response surface methodology techniques. In this study, a response surface methodology based on the Box–Behnken design matrix was used to describe the process of nitrate removal from an aqueous solution with four independent parameters, namely Fe-doped TiO2 (dose 1-2 g l-1), nitrate concentration (25-100 mg l-1), contact time (10-120 min), and pH (4-9). The results indicated that the removal efficiency of nitrate in the presence of ultraviolet and solar radiation was 56.5 % and 21.8%, respectively. The removal efficiency of nitrate increased with time and initial concentration of nitrate. Analysis of variance (ANOVA) indicated that the proposed model was essentially in accordance with the experimental results with the correlation coefficient R2 = 0.9237 and Adj-R2 = 0.8347. Response surface methodology (RSM) proved to be a powerful statistical tool for investigating the operating conditions for nitrate removal under UV irradiation.