Permeability decline occurs during injection of produced water and seawater, resulting in injectivity declines and significant cost increases in waterflooding projects. It is necessary to have a reliable model to predict injectivity decline for preventive water treatment and waterflood design purposes. A classical deep bed filtration (DBF) model has been widely used to predict the injectivity decline. According to this model, the injectivity decline can be characterized by two empirical parameters: filtration coefficient, lambda, and formation damage coefficient; beta. Different methodologies developed to extract these parameters involve expensive and difficult concentration measurements, laboratory-scaled pressure drop measurements (not truly representative of the real reservoir) and simplifying assumptions of analytical solutions. A simple empirical velocity-based damage model proposed by Bachman et al.((1)) is adopted in this work, and extended to multidimensional flow. This model is then compared to the deep bed filtration-based model. The advantage of the empirical model is that it can be easily tuned to either field or laboratory data, and can be easily implemented in reservoir simulators. The paper presents the formulation and numerical implementation of the two coupled reservoir flow and damage models. Different methods of implementing the velocity-based model in multidimensional flow are presented and evaluated. The comparison with the DBF model shows that the two models yield similar damage characteristics. Finally, application of the model to analysis of the published data for offshore Gulf of Mexico is presented. The relationship between the parameters of the two different approaches is validated for these case studies.