The work-up of chemical reactions by standard techniques is often time consuming and energy demanding, especially when chemists have to guarantee low levels of metal contamination in the products. Therefore, scientists need new ideas to rapidly purify reaction mixtures that are both economically and environmentally benign. One intriguing approach is to tether functionalities that are required to perform organic reactions to magnetic nanoparticles, for example, catalysts, reagents, scavengers, or chelators. This strategy allows researchers to quickly separate active agents from reaction mixtures by exploiting the magnetic properties of the support. In this Account, we discuss the main attributes of magnetic supports and describe how we can make the different nanomagnets accessible by surface functionalization. Arguably the most prominent magnetic nanoparticles are superparamagnetic iron oxide nanoparticles (SPIONs) due to their biologically well-accepted constituents, their established size-selective synthesis methods, and their diminished agglomeration (no residual magnetic attraction in the absence of an external magnetic field). However, nanoparticles made of pure metal have a considerably higher magnetization level that is useful in applications where high loadings are needed. A few layers of carbon can efficiently shield such highly reactive metal nanoparticles and, equally important, enable facile covalent functionalization via diazonium chemistry or non-covalent functionalization through π-π interactions. We highlight carbon-coated cobalt (Co/C) and iron (Fe/C) nanoparticles in this Account and compare them to SPIONs stabilized with surfactants or silica shells. The graphene-like coating of these nanoparticles offers only low loadings with functional groups via direct surface modification, and the resulting nanomagnets are prone to agglomeration without effective steric stabilization. To overcome these restrictions and to tune the dispersibility of the magnetic supports in different solvents, we can introduce dendrimers and polymers on Co/C and Fe/C platforms by various synthetic strategies. While dendrimers have the advantage of being able to array all functional groups on the surface, polymers need fewer synthetic steps and higher molecular weight analogues are easily accessible. We present the application of these promising hybrid materials for the extraction of analytes or contaminates from complex aqueous solutions (e.g. waste water treatments or blood analytics), for metal-, organo-, and biocatalysis, and in organic synthesis. In addition, we describe advanced concepts like magnetic protecting groups, a multistep synthesis solely applying magnetic reagents and scavengers, and thermoresponsive self-separating magnetic catalysts. We also discuss the first examples of the use of magnetic scaffolds manipulated by external magnetic fields in flow reactors on the laboratory scale. These hold promise for future applications of magnetic hybrid materials in continuous flow or highly parallelized syntheses with rapid magnetic separation of the applied resins.