The purpose of this study was to clarify the changes in hippocampal glucose metabolism in mild Alzheimer's disease (AD) using positron emission tomography (PET) and 2-(18F)fluoro-2-deoxy-D-glucose (FDG). Forty-one patients with probable mild AD (age: 69.0 +/- 8.0 years; MMSE: 22.6 +/- 2.1) and 22 normal volunteers (age: 67.7 +/- 7.1 years) were studied. The regional cerebral metabolic rate for glucose (CMRglc) was measured using FDG and PET. Although the mean CMRglc in the parietal region was significantly lower in the AD group (right: 6.35 +/- 1.26 mg/100 g/min; left: 6.37 +/- 1.21 mg/100 g/min) than in the control group (right: 7.73 +/- 1.02 mg/100 g/min; left: 7.63 +/- 0.95 mg/100 g/min), the mean CMRglc in the hippocampus did not show a significant difference between the AD group (right: 4.58 +/- 0.70 mg/100 g/min; left: 4.63 +/- 0.67 mg/100 g/min) and the control group (right: 5.22 +/- 0.65 mg/100 g/min; left: 5.22 +/- 0.67 mg/100 g/min) by analysis of variance and post-hoc Tukey's test. The magnitude of the hippocampal CMRglc reduction was not as large as that of parietal CMRglc reduction. Statistical parametric maps (SPM) analysis also did not significantly demonstrate reduced hippocampal CMRglc in AD patients, although it did show a significant reduction in parietal CMRglc in AD patients. Hippocampal CMRglc was not significantly decreased in mild AD. This was unexpected in view of previous studies that have shown atrophy and clinical dysfunction concerning hippocampus in AD, and suggests that the pathophysiology of the hippocampus in AD may be more complex than was previously thought.