Overexpression of the multidrug efflux transport P-glycoprotein may play an important role in pharmacoresistance. 11 C-laniquidar is a newly developed tracer of P-glycoprotein expression. The aim of this study was to develop a pharmacokinetic model for quantification of 11 C-laniquidar uptake and to assess its test-retest variability. Methods: Two (test-retest) dynamic 11 C-laniquidar PET scans were obtained in 8 healthy subjects. Plasma input functions were obtained using online arterial blood sampling with metabolite corrections derived from manual samples. Coregistered T1 MR images were used for region-of-interest definition. Timeactivity curves were analyzed using various plasma input compartmental models. Results: 11 C-laniquidar was metabolized rapidly, with a parent plasma fraction of 50% at 10 min after tracer injection. In addition, the first-pass extraction of 11 C-laniquidar was low. 11 Claniquidar time-activity curves were best fitted to an irreversible single-tissue compartment (1T1K) model using conventional models. Nevertheless, significantly better fits were obtained using 2 parallel single-tissue compartments, one for parent tracer and the other for labeled metabolites (dual-input model). Robust K 1 results were also obtained by fitting the first 5 min of PET data to the 1T1K model, at least when 60-min plasma input data were used. For both models, the test-retest variability of 11 C-laniquidar rate constant for transfer from arterial plasma to tissue (K 1 ) was approximately 19%. Conclusion: The accurate quantification of 11 C-laniquidar kinetics in the brain is hampered by its fast metabolism and the likelihood that labeled metabolites enter the brain. Best fits for the entire 60 min of data were obtained using a dual-input model, accounting for uptake of 11 C-laniquidar and its labeled metabolites. Alternatively, K 1 could be obtained from a 5-min scan using a standard 1T1K model. In both cases, the test-retest variability of K 1 was approximately 19%.