This retrospective audit was conducted to examine the changes in patient characteristics, referral, treatment and outcome over a 20-year period in a large regional neuro-oncology centre, focusing on the impact of the changes in pathological classification of gliomas. Using the Edinburgh Cancer Centre (ECC) database all cases of glioma were identified and patient, tumour and treatment characteristics noted. Survival was calculated from date of surgery or, if no operation was performed, the date of referral. Comparison was made between four periods 1988-1992 (c1), 1993-1997(c2), 1998-2002(c3) and 2003-2007 (c4). During the 20 years, 1175 patients with a glioma were referred to ECC. The median age increased from 53 years to 57 years (p < 0.001) but the proportion without pathology remained unchanged (10%). The distribution of pathological grades changed over time Grade I-II: 24, 6, 6, and 6%, Grade III: 42, 27, 17, and 13% and Grade IV: 24, 61, 68, and 68% in c1, c2, c3 and c4, respectively (p < 0.001). Immediate RT was given to 68% (c1), 70% (c2), 78% (c3) and 79% (c4). Median interval from resection to RT reduced from 43 days (c1) to 36 days (c4) (p < 0.001). 5-year overall survival for patients with Grade III lesions increased: 21% (c1), 35% (c2), 37% (c3), 33% (c4) as did 1-year overall survival for Grade IV lesions: 18% (c1), 26% (c2), 29% (c3), 27% (c4)). This improvement probably reflects the change in pathological classification rather than a change in management. Proportional hazards analysis of grade IV 1993-2007 only (to reduce pathological variation) showed that younger age, frontal lesions, excision, higher RT dose had reduced hazard of death. Interval from surgery to RT had no impact on survival in this series.