The three-dimensional (3D) structure of chromosomes is associated with gene regulation and cell function. Understanding how 3D genome structure varies between cell types and states, in development and disease, promises to enhance the interpretation of genome sequence variation and to accelerate the discovery of disease target genes. To explore 3D genome structure variation in early development and highly specialised cells in the brain, we applied Genome Architecture Mapping in different cell types from the adult murine brain: oligodendroglia (OLGs) from the cortex, dopaminergic neurons (DNs) from the midbrain and pyramidal glutamatergic neurons (PGNs) from the hippocampus (Winick-Ng et al. 2021). We found extensive cell-type specialisation of 3D chromatin contacts, such as extensive reorganisation of topological domains (TADs) and eu/heterochromatic compartments. We also discover the large scale decondensation, or ‘melting’, of long genes when most highly expressed, many with roles in neurodevelopmental and neurodegeneration disorders. Through integration of 3D genome structure with single-cell expression and chromatin accessibility, we find that PGN- and DN-specific hubs of contacts containing genes associated with their specialised functions, such as addiction and synaptic plasticity, respectively. Our current work explores advantages of GAM relative to ligation-based conformation capture methods (Beagrie, Thieme, et al. 2022), differences in 3D genome structure between parental chromosomes (Markowski et al. 2021), how genetic variation associated with disease interferes with 3D genome structure (Yanchus et al. 2022), and the effects of environmental insults on the complex 3D genome structures of brain cells.