Fluorescence imaging of immunolabeled brain slices is a key tool in neuroscience that enable mapping of proteins or DNA/RNA at resolutions not possible with non-invasive techniques, including magnetic resonance or nuclear imaging. The signal in specific regions is usually quantified after manually drawing regions of interest, risking operator-bias. Automated segmentation methods avoid this risk but require multi-sample average atlases with similar image contrast as the images to be analyzed. We here present the first population-based average atlas of the C57BL/6 mouse brain constructed from brain sections labeled with the fluorescence nuclear stain DAPI. The data set constitutes a rich three-dimensional representation of the average mouse brain in the DAPI staining modality reconstructed from coronal slices and includes an automatic segmentation/spatial normalization pipeline for novel coronal slices. It constitutes the final population-based average template, individual reconstructed brain volumes, and native coronal slices. The comprehensive data set and accompanying spatial normalization/segmentation software are provided. We encourage the community to utilize it to improve and validate methods for automated brain slice analysis.