In the last decade, super-resolution microscopy has evolved as a very powerful method for sub-diffraction resolution fluorescence imaging of cells and structural investigations of cellular organelles. Super-resolution microscopy methods can now provide a spatial resolution that is well below the diffraction limit of light microscopy, enabling invaluable insights into the spatial organization of proteins in biological samples. However, current super-resolution measurements become error-prone below 25 nm. An alternative approach to bypass the diffraction limit and enable “super-resolution imaging” on standard fluorescence microscopes, is the physical expansion of the cellular structure of interest. By linking a protein of interest into a dense, cross-linked network of a swellable polyelectrolyte hydrogel, biological specimens can be physically expanded allowing ~70 nm lateral resolution by confocal laser scanning microscopy. Since its first introduction by Boyden and co-workers in 2015, expansion microscopy (ExM) has shown impressive results including the magnified visualization of pre- or post-expansion labeled proteins and RNAs with fluorescent proteins, antibodies, and oligonucleotides, respectively, in cells, tissues, and human clinical specimen. By combining ExM with single-molecule localization microscopy (SMLM) it is potentially possible to approach the resolution of electron microscopy. However, current attempts to combine both methods remained challenging because of protein and fluorophore loss during digestion or denaturation, gelation, and the incompatibility of expanded polyelectrolyte hydrogels with photoswitching buffers. Here we show that re-embedding of expanded hydrogels enables dSTORM imaging of expanded samples and demonstrate that post-labeling ExM resolves the current limitations of super-resolution microscopy. Using reference structures, neurons and brain slices, we demonstrate that post-labeling Ex-SMLM can be used advantageously for super-resolution imaging. It preserves ultrastructural details, improves the labeling efficiency and reduces the positional error arising from linking fluorophores into the gel thus paving the way for super-resolution imaging of immunolabeled endogenous proteins with true molecular resolution.