Factors contributing to the quantification of the striatal dopamine D2 receptor binding potential in vivo using I-123-iodobenzamide (I-123-IBZM) and SPECT were analyzed in phantom studies, healthy volunteers and in patients with the parkinsonian syndrome. A cylindrical brain phantom based on a stereotactic brain atlas was constructed with independently fillable compartments representing two striata (ST), cerebellum (CB) and background. Clinical I-123-IBZM SPECT studies were performed on 15 healthy volunteers and on 28 patients with parkinsonian syndrome. Interobserver variability of region of interest (ROI) selection and count ratios were estimated by two independent observers. ROIs for the striatum were either fixed, based on a stereotactic brain atlas, or drawn manually, based on 70% isocontour lines. Reference regions were either the cerebellum (isocontour ROIs) or the occipital cortex (occiptal cortex; fixed ROIs). The brain phantom measurements showed linearity with respect to radioactivity concentration, good reproducibility and good contrast recovery. The interobserver study-showed that the striatum-to-occiptal cortex ratio with fixed ROIs for the striatum, as an estimate for striatal D2 receptor binding potential, resulted in a means of separating patients with normal receptor activity from those with decreased striatal dopamine D2 receptor activity.