The purpose of this study was to assess the utility of a new single-photon emission tomography ligand, [123I]5-iodo-3-[2(S)-2-azetidinylmethoxy]pyridine (5-I-A-85380), to measure regional nAChR binding in human brain. Six healthy nonsmoker subjects (two men and four women, age 33 +/- 15 years) participated in both a bolus (dose: 317 +/- 42 MBq) and a bolus plus constant infusion (dose of bolus: 98 +/- 32 MBq, B/I=6.7 +/- 2.6 h, total dose: 331 +/- 55 MBq) study. The study duration was 5-8 h and 14 h in the former and the latter, respectively. Nonlinear least-squares compartmental analysis was applied to bolus studies to calculate total (VT') and specific (VS') distribution volumes. A two-tissue compartment model was applied to identify VS'. VT' was also calculated in B/I studies. In bolus studies, VT' was well identified by both one- and two-tissue compartment models, with a coefficient of variation of less than 5% in most regions. The two-compartment model gave VT' values of 51, 22, 27, 32, 20, 19, 20, and 17 ml cm(-3) in thalamus, cerebellum, putamen, pons, and frontal, parietal, temporal, and occipital cortices, respectively. The two-compartment model did not identify VS' well. B/I studies provided poor accuracy of VT' measurement, possibly due to deviations from equilibrium conditions. These results demonstrate the feasibility of quantifying high-affinity type nAChRs using [123I]5-I-A-85380 in humans and support the use of VT' measured by bolus studies.