Magnetic particles, generally nanostructured and magnetite-based, have been used extensively to remove drinking water contaminants via adsorption or catalytic degradation. Compositions alternative to Fe3O4 could address long-standing issues of magnetic recoverability and material integrity in real waters. Two alternative compositions of magnetic nanoparticles (Fe3C@C and AuFe) are studied as nano-absorbents and nano-catalysts, respectively. The stability, magnetic separability, and adsorptive properties of nanostructured carbon-coated iron carbide (Fe3C@C) were compared to those of Fe3O4 and other common iron oxide-based nanomaterials. Experimental results show that (i) Fe3C@C is chemically stable in simulated drinking water, (ii) can be separated from water magnetically under flow with >99% recovery, and (iii) is capable of removing organics (1.60 mg-MB/m2) or oxo-anions (6.75 µg-As/m2) from simulated drinking water. In terms of scale up, an optimized permanent magnetic nanoparticle recovery device (i.e., the MagNERD) was developed and operated to separate, capture and reuse superparamagnetic nanoparticles from treated water in-line under continuous flow conditions. Experimental data and computational modeling demonstrate how the MagNERD’s efficiency to recover nanoparticles depends upon reactor configuration, including the integration of stainless-steel wool around permanent magnets, hydraulic flow conditions, and magnetic nanoparticle uptake. The MagNERD efficiently removes high concentrations (500 ppm) of magnetic nanopowder (e.g., >95% removal) under scalable and process-relevant flow rates (e.g., 1 L/min through a 1.11-L MagNERD reactor) from varying water matrices (e.g., ultrapure water, brackish water) and after treatment of As contaminated simulated drinking water (e.g., >94% removal of arsenic-bound Fe3O4). Additionally, we investigate nano-magnetism for the purposes other than magnetic removal – magnetic nanoparticle heating – and explore the structure property relationship between a classically non-magnetic material (e.g., nano-hematite) and magnetic nanoparticle heating. Lastly, the legitimacy of using a magnetic and environmentally bengin AuFe bimetallic catalyst for environmental remediation is explored. Using nitroarene reduction – a simple model environmental reaction – as a probe reaction and AuFe bimetallic catalyst, we found applying an alternating magnetic field (AMF) increased nitroarenes reduction reaction rates by 200%, via localized particle surface heating. This rate constant was equivalent to a ~27C bulk reaction temperature, suggesting AMF exposure raised the AuFe NP surface temperature ~4C above ambient.