Performing quantitative small-animal PET with an arterial input function has been considered technically challenging. Here, we introduce a catheterization procedure that keeps a rat physiologically stable for 1.5 mo. We demonstrated the feasibility of quantitative small-animal 18 F-FDG PET in rats by performing it repeatedly to monitor the time course of variations in the cerebral metabolic rate of glucose (CMR glc ). Methods: Aseptic surgery was performed on 2 rats. Each rat underwent catheterization of the right femoral artery and left femoral vein. The catheters were sealed with microinjection ports and then implanted subcutaneously. Over the next 3 wk, each rat underwent 18 F-FDG quantitative small-animal PET 6 times. The CMR glc of each brain region was calculated using a 3-compartment model and an operational equation that included a k* 4 . Results: On 6 mornings, we completed 12 18 F-FDG quantitative small-animal PET studies on 2 rats. The rats grew steadily before and after the 6 quantitative small-animal PET studies. The CMR glc of the conscious brain (e.g., right parietal region, 99.6 6 10.2 mmol/100 g/min; n 5 6) was comparable to that for 14 C-deoxyglucose autoradiographic methods. Conclusion: Maintaining good blood patency in catheterized rats is not difficult. Longitudinal quantitative small-animal PET imaging with an arterial input function can be performed routinely. Sever al small-animal models have been developed that simulate many of the important hallmarks of human brain diseases. Although the first report of quantitative small-animal PET to monitor the CMR glc changes in a rat model of brain injury appeared in 2000 (1), there has been rare use of quantitative small-animal PET with experimental models over the intervening years. Cannulation of a rat is not difficult. The challenges associated with repeated surgeries to obtain arterial blood samples have most likely impeded routine use of the method. Although image-derived input function appears to be an attractive alternative to arterial sampling, imagederived input function is reliable only in selected situations and selected tracers (2). In the case of 18 F-FDG quantitative smallanimal PET, one of the major concerns was the different 18 F-FDG concentrations in plasma and in whole blood (WB) due to a slow erythrocyte 18 F-FDG transport rate in rodent blood (3,4). Therefore, arterial plasma samples remain the gold standard for input function and CMR glc quantification in rats (5).To ease blood sampling and reduce blood loss in small animals, we developed an automated microfluidic device that allows us to take discrete, serial blood samples (,1 mL per sample) with little human effort (6). However, blood coagulation took place frequently in the catheter when the interval between 2 serial blood samples was long (e.g., .10 min). In this work, we developed a surgical procedure that eases radiotracer injection and arterial blood sampling. We conducted a proof-of-principle study to demonstrate the feasibility of performing multiple-time-point quanti...