A two-step magnetically seeded filtration process that includes heterogeneous flocculation (shear-flow and Brownian) and magnetic filtration is examined experimentally. The effects of various parameters - magnetic-field strength, size of particles, flow rate, seeding concentration? and solution pH -on the removal efficiency are investigated. A breakthrough model - which combines trajectory analysis, a particle buildup model, and a bivariate population-balance model applicable for Brownian flocculation - is developed to predict particle breakthrough in a magnetic filter. Experiments show that the removal efficiency increases as magnetic-field strength and particle size are increased and flow rate is decreased. A maximum in the removal efficiency is observed at a certain seeding concentration and at the lower pH values, which is explained from competing effects that take place with respect to magnetic susceptibility and size of aggregates as the seeding concentration and solution pH are increased. Modeling results of the trajectory analysis show that the effect of hydrodynamic resistance becomes important as Reynolds number and particle size are increased or the magnetic-field strength is decreased. Similarly to experimental observations, the modeling results predict that the removal efficiency increases with increasing magnetic-field strength and particle size indicating that the relative importance of magnetic and drag forces and the aggregation rate in the flocculation step play an important role in the magnetically seeded process. The breakthrough model developed in this study provides a good description of the experimental breakthrough data obtained from magnetic filtration of paramagnetic particles and magnetically seeded filtration with Brownian-flocculation.