The in vitro utilization of biocompatible ferromagnetic nanoparticles (BFNs) in hemodialysis (HD), routinely used today for the treatment of end stage renal disease (ESRD), is introduced in this work. The proposed strategy is termed magnetically assisted hemodialysis (MAHD) and it aims to become a more efficient development of conventional HD. The method is based on the production of biocompatible ferromagnetic nanoparticles-targeted binding substances conjugates (BFNs-TBSs Cs) constructed of BFNs and specifically designed TBSs that should have high affinity and binding capacity for target toxic substances (TTSs) which must be removed from the ESRD patient subjected to HD. Antibodies or even specific proteins could serve as the TBS of the desired BFNs-TBSs Cs. The BFNs-TBSs Cs should be administered to the patient timely prior to the MAHD session so as to bind with the desired TTSs during their free circulation in the vascular network. Eventually, the complete BFNs-TBSs-TTSs structure can be selectively removed during the MAHD session by means of an external inhomogeneous magnetic field that is applied either at the dialyzer or at other collection point(s) along the blood circulation line of the dialysis machine. The advantages of MAHD over conventional HD regarding the patient's comfort and overall health status are discussed in detail among practical issues. To examine this proposition we employed Fe(3)O(4) and bovine serum albumin (BSA) as the BFN and the TBS constituents respectively, since they are both highly biocompatible. By means of x-ray diffraction, atomic force microscopy, circular dichroism spectropolarimetry, UV-vis spectrophotometry, SQUID magnetometry, and nuclear magnetic resonance we evaluated (i) the structural/morphological characteristics, (ii) the magnetic retraction efficiency, and most importantly (iii) the toxin binding affinity and capacity of both bare Fe(3)O(4) BFNs and Fe(3)O(4)-BSA Cs by performing in vitro experiments on specific TTSs. Homocysteine and p-cresol were chosen as representative TTSs and were investigated in great detail. The results obtained prove the in vitro applicability of the proposed MAHD method. Corrections were made to this article on 6 November 2007 (see figure 10 caption and lines 10 and 11 of page 11). The corrected electronic version is identical to the print version.