For clinical dynamic contrast-enhanced (DCE) MRI studies, it is often not possible to obtain reliable arterial input function (AIF) in each measurement. Thus, it is important to find a representative AIF for pharmacokinetic modeling of DCE-MRI data when individual AIF (Ind-AIF) measurements are not available. A total of 16 patients with osteosarcomas in the lower extremity (knee region) underwent multislice DCE-MRI. Reliable Ind-AIFs were obtained in five patients with a contrast injection rate of 2 cc/s and another five patients with a 1 cc/s injection rate. Average AIF (Avg-AIF) for each injection rate was constructed from the corresponding five Ind-AIFs. For each injection rate there are no statistically significant differences between pharmacokinetic parameters of the five patients derived with Ind-AIFs and Avg-AIF. There are no statistically significant changes in pharmacokinetic parameters of the 16 patients when the two AvgAIFs were applied in kinetic modeling. The results suggest that it is feasible, as well as practical, to use a limited-population- There has been increasing interest in the T 1 -weighted dynamic contrast-enhanced (DCE) MRI method for the study of many different tumor types, using the Gd (III) chelate contrast agents (1). There are generally three approaches for analyzing DCE-MRI signal time courses: 1) qualitative subjective assessment of curve shape, such as wash-out, plateau, and persistent; 2) empirical quantitation, such as maximum slope and percent signal intensity change; and 3) analytical pharmacokinetic modeling. The latter is more sophisticated, and also the more desirable. Unlike the first two approaches, analytical modeling of DCE-MRI data extracts pharmacokinetic parameters that should be independent of data acquisition details, contrast agent dose and injection rate, magnetic field strength, and vendor platform, etc. This improves DCE-MRI study reproducibility and enables meaningful comparison of results from different imaging sites where different DCE-MRI protocols are employed. The extracted pharmacokinetic parameters are usually variants of: K trans , a rate constant for contrast agent plasma/interstitium transfer, and v e , the interstitial space volume fraction (the putative contrast agent distribution volume). These parameters have been used for cancer diagnosis (2-4) and monitoring effects of antiangiogenic therapies (5,6).A characteristic aspect of pharmacokinetic modeling of DCE-MRI signal time course is the requirement for an arterial input function (AIF). The absolute accuracy of the pharmacokinetic parameters, K trans and v e , depends on the AIF accuracy (7,8). Ideally, individually measured AIF should be used for kinetic modeling of the corresponding tissue DCE-MRI data in each experiment (3,7,9). However, for patient studies that are conducted in clinical settings, it is often not possible to obtain reliable AIF measurement in each DCE-MRI examination, due to either data acquisition constraints, such as excitation volume coverage and image slice angulation, o...