In Study 1, the effect of acute hypoglycaemia on memory consolidation was studied in non-diabetic subjects. Functional brain imaging was used to identify regions with altered metabolism following hypoglycaemia. No effect of hypoglycaemia was seen on memory or brain imaging results, suggesting that acute hypoglycaemia does not affect previously-formed memories, though other possibilities are discussed. In Study 2, the effects of acute hypoglycaemia on learning and recall were studied, and a novel test of prospective memory, mimicking memory in real life, was developed. Subjects had type 1 diabetes and either normal or impaired awareness of hypoglycaemia. Both learning and recall were impaired during hypoglycaemia. Prospective memory was impaired to a similar degree, reinforcing the ecological validity of laboratory cognitive function testing. No statistically significant differences between normal- and impaired-awareness subjects were seen, but this may have been due to lack of statistical power. A previous study had unexpectedly reported that high-level cognitive functions were unaffected by hypoglycaemia. Study 3 re-addressed this issue using a more difficult cognitive test to exclude the possibility of a ceiling effect, and impairment by hypoglycaemia was confirmed. A separate task, intended to indicate which of speed and accuracy of cognitive processing is primarily affected, yielded no significant results. In study 4, the symptoms of hyperglycaemia reported by 400 people with insulin-treated diabetes were examined using factor analysis. Four categories were identified, and labelled as ‘osmotic’, ‘neurological’, ‘mental agitation’ and ‘malaise’. Substantial overlap with hypoglycaemia symptoms was noted. The mean blood glucose level for symptom onset was 15 mmol/l, with a tendency for less intense symptoms and a higher symptom threshold in older people and those with impaired hypoglycaemia awareness.