[11C]FLB 457 is a radioligand for positron emission tomography (PET) that possesses high affinity to D2/D3 receptors. It has been suggested to be useful for quantification of low-density dopamine D2 receptor populations, e.g. in cortical and limbic brain areas. We explored the reproducibility of five methods for measuring extrastriatal D2-like receptor binding potential with [11C]FLB 457. Seven healthy male volunteers were examined twice with [11C]FLB 457 (high specific radioactivity) on the same day, at least 3 h apart. Four brain areas, frontal cortex, nucleus thalamus, temporal cortex and cerebellar cortex, were examined. Binding potentials (BPs) were derived from (1) a target to cerebellum distribution volume ratio, (2/3) two reversible reference tissue compartment models and (4) a transient equilibrium approach. For comparison, BP values were also calculated with the standard three-compartment kinetic model that does not assume a receptor-free reference region. The use of the standard three-compartment model did not result in reproducible BP estimates. The distribution volume (DV) ratio, reference tissue compartment models and the transient equilibrium method all had good to excellent intraclass correlation coefficients (ICCs) in the studied brain areas ranging from 0.56 to 0.93. Absolute variability was also relatively low, ranging from 5.3% to 10.4%. There were no marked differences in the ICC or absolute and relative variability between the four methods based on a reference tissue (cerebellum). In addition, we did not observe systematic differences in the BP between the first and the second scan. These data indicate that the reproducibility of the DV ratio, reference tissue models and the transient equilibrium method is good or excellent. However, each of these methods includes assumptions affecting their validity. Thus, the choice of method will be critically dependent on the purpose of the study.