This study proposes a practical approach, using the minimum number of brain regions and stains, to consolidate previously published neuropathological criteria into one operationalised schema to differentiate subtypes of frontotemporal lobar degeneration with tau-immunopositive inclusions (FTLD-tau). This approach uses the superior frontal and precentral cortices and hippocampus stained for phosphorylated-tau, p62 and modified Bielschowsky silver, and the midbrain stained only for modified Bielschowsky silver. Accuracy of interrater reliability was determined by ten raters in 24 FTLD-tau cases (Pick’s disease = 4, corticobasal degeneration = 9, progressive supranuclear palsy = 5, globular glial tauopathy = 6) including four with a mutation in MAPT collected with consent by Sydney Brain Bank. All brain regions and stains assessed proved informative for accurate pathological subtyping, and many neuropathological features were identified as common across the FTLD-tau subtypes. By identifying subtype-specific neuropathological features in the sections selected, ten independent observers assigned the cases to a FTLD-tau subtype with almost perfect agreement between raters, emphasising the requirement for the assessment of subtype-specific features for the accurate subtyping of FTLD-tau. This study consolidates current consensus diagnostic criteria for classifying FTLD-tau subtypes with an efficient, simple and accurate approach that can be implemented in future clinicopathological studies.