We have studied the progression of striatal and extrastriatal post-synaptic dopaminergic changes in a group of 12 patients with Huntington's disease using serial (11)C-raclopride PET, a specific marker of D2 dopamine receptor binding. All patients had two (11)C-raclopride PET scans 29.2 +/- 12.8 months apart, and six of them had a third scan 13.2 +/- 3.9 months later. We found a mean annual 4.8% loss of striatal (11)C-raclopride binding potential (BP) between the first and second scans, and a 5.2% loss between the second and third scans. Statistical Parametric Mapping (SPM) localized significant baseline reductions in (11)C-raclopride BP in both striatal and extrastriatal areas, including amygdala, temporal and frontal cortex in Huntington's disease compared with normal subjects matched for age and sex. When the (11)C-raclopride scans performed 29 months after the baseline scans were considered, SPM revealed further significant striatal, frontal and temporal reductions in (11)C-raclopride BP in Huntington's disease. Cross-sectional Unified Huntington's Disease Rating Scale (UHDRS) scores correlated with (11)C-raclopride binding, but there was no correlation between individual changes in UHDRS motor scores and changes in striatal binding. Performance on all neuropsychological measures deteriorated with time but only the accuracy score of the one-touch Tower of London test correlated significantly with striatal and putamen D2 binding. In summary, serial (11)C-raclopride PET demonstrates a linear progression of striatal loss of D2 receptors in early clinically affected Huntington's disease patients over 3 years. SPM also revealed a progressive loss of temporal and frontal D2 binding. Changes over time in clinical scores and in neuropsychological assessments, except for measures of planning, did not correlate with striatal D2 binding. This probably reflects both contributions from other affected brain structures and high variance in these measures.