Abstract In order to fully understand the structural substrates underlying the brain function, it is central to curate multiple attributes of the same neurons. The important attributes include, but limited to morphology, connection properties and molecular patterns, such as the expression of functional genes and the distribution of synaptic densities. Light microscopy-based neuronal tracing has contributed our fundamental understanding of the heterogeneity of neuronal morphology. Aligning morphology reconstructions of extremely sparsely labeled neurons from multiple brain samples onto a common coordinate system permitted systematic comparison of the receptive dendritic fields and the projective axonal arbors of distinct neuronal subtypes. However, connection partners of neurons cannot be identified from the light microscopy-based reconstructions due to the nature of sparse sampling and limited resolution. On the other hand, electron microscopy-based reconstructions label all the cell nondifferential but permit recognition of ultra-fine structures, including synaptic contacts, which in theory permits mapping the complete connectome between all neurons. However, it is technically and economically challenging to scale lossless electron microscopy to a large sample, such as the mouse brain. In addition, molecular information is often lost in sample processing. In this project, we propose to develop a scalable strategy that combines several novel technologies to permit reconstructing a molecular connectome of the mammalian brain by light microscopy. Being able to simultaneous profile neuronal morphology, connectivity and molecular properties throughout the whole mouse brain will clarify our understanding of the heterogeneity of brain cells and advance our knowledge about the overall architecture of the mammalian brain with unprecedented resolution and scale.