Traditional histopathological evaluation of gliomas relies on haematoxylin and eosin (HE) staining of sectioned tissue. Whilst this technique allows excellent visualisation of nuclei and cytoplasm alike, there are a number of limitations. For example, only a small number of sections from each case can be analysed meaning that key diagnostic features may be missed. To circumvent this problem we aimed to develop a novel technique for 3D-imaging of large glioblastoma specimens with HE-like labelling. We identified cresyl violet as a suitable fluorescent haematoxylin analogue for nuclei labelling in glioblastoma. This was followed by rapid tissue clearing using a pathoDISCO approach. Subsequently, samples were imaged with high resolution by light sheet microscopy; two stacks were recorded – one exhibiting nuclear labelling by cresyl violet, the other recording autofluorescence to serve as an eosin analogue. Subsequently, stacks were processed using a false-colouring algorithm to produce virtual HE images, which were then volumetrically rendered. This approach allowed us to visualise glioblastoma samples in 3D with HE-like labelling in a comparable timeframe to standard HE preparations. Various hallmark histological findings of glioblastoma were identified in our virtual HE images, indicating accurate recapitulation of staining patterns seen in standard HE preparations. In addition, we were able to visualise glioblastoma vasculature with a high level of detail. This study represents an exciting and novel approach to 3D histopathology of tumour specimens. 3D virtual histopathology has the potential to revolutionise and improve the way cancers are diagnosed, possibly leading to better informed treatment regimens and patient outcomes.