The radiolabeled serotonin transporter (SERT) ligand [(11)C](+)-McN5652 has recently been used in clinical positron emission tomography (PET) studies for SERT imaging. However, this radioligand offers disadvantages in routine clinical settings because of its short radioisotope half-life (eg PET facilities within hospitals without a cyclotron need to acquire such radioligands from distant cyclotron units for clinical use). S-([(18)F]fluoromethyl)-(+)-McN5652 ([(18)F](+)-FMe-McN5652) is an analogue which has been synthesized newly, and has a significantly longer radioisotope half-life. In the porcine brain, it demonstrates the same characteristic distribution pattern of serotonin-uptake sites like the (11)C-labeled congener with the highest binding in the midbrain and thalamus and the lowest in the cerebellum and occipital cortex. It shows a 30% higher blood-brain transfer and a slower peripheral metabolism than [(11)C](+)-McN5652. Rather uniform brain binding was observed after injection of the pharmacologically inactive radiolabeled enantiomer, or after pretreatment with the highly selective SERT inhibitor citalopram. The norepinephrine uptake inhibitor maprotiline did not show any inhibitory effect. Using a one-tissue compartment model (K(1), k"(2)) or a two-tissue compartment model (K(1) to k(4)) with or without constraints for calculation, the regional binding parameters of [(11)C](+)-McN5652 and [(18)F](+)-FMe-McN5652 are highly correlated among each other and with the SERT density, as determined by in vitro binding of [(3)H]citalopram. Using constraints to correct for the free fraction and nonspecific binding of the radiotracers, a considerable increase of the midbrain-occipital cortex ratios with higher values for [(18)F](+)-FMe-McN5652 compared to [(11)C](+)McN5652 was revealed. It is concluded that [(18)F](+)-FMe-McN5652 has better features than [(11)C](+)McN5652 for SERT imaging with PET.