Corticobasal syndrome, progressive supranuclear palsy and frontotemporal dementia are all part of a disease spectrum that includes common cognitive impairment and movement disorders. The aim of this study was to characterize brain cholinergic deficits in these disorders. We measured brain acetylcholinesterase activity by [11C] N-methylpiperidin-4-yl acetate and positron emission tomography in seven patients with corticobasal syndrome (67.6+/-5.9 years), 12 with progressive supranuclear palsy (68.5+/-4.1 years), eight with frontotemporal dementia (59.8+/-6.9 years) and 16 healthy controls (61.2+/-8.5 years). Two-tissue compartment three-parameter model and non-linear least squares analysis with arterial input function were performed. k3 value, an index of acetylcholinesterase activity, was calculated voxel-by-voxel in the brain of each subject. The k3 images in each disease group were compared with the control group by using Statistical Parametric Mapping 2. Volume of interest analysis was performed on spatially normalized k3 images. The corticobasal syndrome group showed decreased acetylcholinesterase activity (k3 values) in the paracentral region, frontal, parietal and occipital cortices (P<0.05, cluster corrected). The group with progressive supranuclear palsy had reduced acetylcholinesterase activity in the paracentral region and thalamus (P<0.05, cluster corrected). The frontotemporal dementia group showed no significant differences in acetylcholinesterase activity. Volume of interest analysis showed mean cortical acetylcholinesterase activity to be reduced by 17.5% in corticobasal syndrome (P<0.001), 9.4% in progressive supranuclear palsy (P<0.05) and 4.4% in frontotemporal dementia (non-significant), when compared with the control group. Thalamic acetylcholinesterase activity was reduced by 6.4% in corticobasal syndrome (non-significant), 24.0% in progressive supranuclear palsy (P<0.03) and increased by 3.3% in frontotemporal dementia (non-significant). Both corticobasal syndrome and progressive supranuclear palsy showed brain cholinergic deficits, but their distribution differed somewhat. Significant brain cholinergic deficits were not seen in frontotemporal dementia, which may explain the unresponsiveness of this condition to cholinergic modulation therapy.