Macroscopic modeling approaches based on the solution of the Richards equation with root water uptake (RWU) as a sink term can help in understanding soil-water-plant interactions within the rhizosphere. However, these models currently cannot capture the differences in RWU attributed to variations in plant health. Errors in simulating RWU from unhealthy plants are significant when disease-causing fungus inhibits water uptake rather than other usually considered plant stresses. We developed RWU reduction functions to simulate plant transpiration under combined water and disease stress conditions using linear and non-linear response models. The developed functions were implemented in the numerical model HYDRUS (2D/3D) to simulate water uptake from a root system in a radially symmetrical flow domain. Field experiments were conducted in the Vidarbha region of central India for one crop cycle on four citrus trees with varying disease intensities (healthy to severely diseased). The proposed model was rigorously tested by comparing its results with measured soil water contents and plant transpiration fluxes under various water and disease limiting conditions. Error in simulating RWU fluxes from unhealthy trees by ignoring the disease stress factor was found to be significant (15% for slightly diseased to 26% for the severely diseased tree). Parameters of the spatial root distribution and the disease stress response functions were optimized for each scenario using a genetic algorithm approach. Our results indicate that calibration targets to validate uptake reduction functions should be chosen cautiously based on the dominant stress experienced by the plant root system. © 2020 Elsevier B.V.