In this paper we describe the magnetorheological (MR) behavior of aqueous suspensions consisting of magnetite particles stabilized by poly(acrylic acid) polymers (PAA). A previous work on the colloidal stability of the same systems for different pH values and polymer concentrations demonstrated that the addition of PAA polymers has a very significant effect on the stability. In the present contribution, we study the MR effect of the suspensions stabilized by two different commercial polymers, as a function of pH, magnetic field strength and magnetite volume fraction. All the results are discussed in terms of the interfacial properties of the systems. It is demonstrated that for a given concentration of micrometer particles, the theological response strongly depends on pH, on the volume fraction of magnetite particles, on the type of polymer added for increasing the stability and on the magnetic field strength. Changing the polymer used provokes clear rheological differences for the same sample conditions (field strength, volume fraction and pH). This is suggested to be due to the hydrophobic/hydrophilic balance of the polymer affecting the magnetic field ability to form magnetic structures by aggregation of the magnetized particles. The results are compared to the predictions of the so-called standard chain model, based on the assumption that the MR effect is the result of the balance between the magnetic interactions (tending to establish some degree of order in the suspension by formation of particle chains in the direction of the field) and hydrodynamic ones (tending to destroy the formed structures by viscous stress on the chains). it is found that the behavior of the yield stress does not agree well with the predictions of the model when the relative proportion of both particle and polymer confers optimum stability to the dispersions. This is likely due to the fact that the presence of the stabilizing polyelectrolyte provokes that the magnetic field is not as effective in structuring the suspension as deduced from the chain model. (c) 2008 Elsevier Inc. All rights reserved.