Aluminium, iron and copper are all implicated in the aetiology of neurodegenerative diseases including Alzheimer's disease. However, there are very few large cohort studies of the content of these metals in aged human brains. We have used microwave digestion and TH GFAAS to measure aluminium, iron and copper in the temporal, frontal, occipital and parietal lobes of 60 brains donated to the Cognitive Function and Ageing Study. Every precaution was taken to reduce contamination of samples and acid digests to a minimum. Actual contamination was estimated by preparing a large number of (170+) method blanks which were interspersed within the full set of 700+ tissue digests. Subtraction of method blank values (MBV) from tissue digest values resulted in metal contents in all tissues in the range, MBV to 33 μg g(-1) dry wt. for aluminium, 112 to 8305 μg g(-1) dry wt. for iron and MBV to 384 μg g(-1) dry wt. for copper. While the median aluminium content for all tissues was 1.02 μg g(-1) dry wt. it was informative that 41 brains out of 60 included at least one tissue with an aluminium content which could be considered as potentially pathological (> 3.50 μg g(-1) dry wt.). The median content for iron was 286.16 μg g(-1) dry wt. and overall tissue iron contents were generally high which possibly reflected increased brain iron in ageing and in neurodegenerative disease. The median content for copper was 17.41 μg g(-1) dry wt. and overall tissue copper contents were lower than expected for aged brains but they were commensurate with aged brains showing signs of neurodegenerative disease. In this study we have shown, in particular, the value of carrying out significant numbers of method blanks to identify unknown sources of contamination. When these values are subtracted from tissue digest values the absolute metal contents could be considered as conservative and yet they may still reflect aspects of ageing and neurodegenerative disease in individual brains.