Perfusion measures the level of blood supply to body tissues, and its abnormality conveys important physiological information for diagnosis, especially for cancer. Comparing to conventional contrast based perfusion, arterial spin labeling (ASL) based perfusion is contrast free and offers quantification without the need to time the arrival of contrast bolus. Although growing popular in the neuroimaging, the use ASL based perfusion in body remains largely in the research field, mostly constrained by the susceptibility related image distortion since single shot acquisition is generally used. A straightforward way to reduce these effects is to shorten the readout train length, which however leads to lowered spatial resolution. In this work, we investigate the use of reduced FOV excitation [1] in ASL based perfusion to shorten the readout train length while maintaining the spatial resolution, in the application of prostate cancer.Flow sensitive alternating inversion recovery (FAIR) [2] based pulsed ASL perfusion with rFOV excitation was implemented as illustrated in Fig.1. In FAIR perfusion module, a hyperbolic scant pulse was used to achieve selective and global inversion. A Type-I 2D RF [1] was used for excitation with spatial selection in both phase encoding and slice selective direction. Single shot spin echo EPI (SE-EPI) was used for readout. All MR experiments were performed in a 3.0T whole body system (GE Discovery 750). A patient diagnosed with prostate cancer underwent MR examination including multi-phase FAIR perfusion acquisition and multi-phase DCE liver acceleration volume acquisition (LAVA). Consent form was acquired prior to the study. The imaging parameters for FAIR were following: FAIR, FOV 80 * 48 mm; slice thickness 10 mm; matrix 48 * 48. Inversion time (TI) of 1.0, 1.2, 1.4 s were used with recovery time of 3.0 s. Selective inversion slab with a thickness of 20 mm was used to uniformly invert imaging plane. Sixteen pairs of label and control acquisitions were made for averaging. LAVA used the following parameters: TE/TR = 1.3/3.0 ms; FOV 360 * 360 mm; slice thickness 3.0 mm; slice number 64; matrix 256*160; and flip angle 15^o. Forty phases were acquired with temporal resolution of 9.0 s. DCE pharmacokinetics was analyzed in GEN-IQ (GE Advanced workstation 4.6).T2, T2 fused K^trans map and T2 fused rFOV ASL perfusion map (TI = 1.2 s) are shown in Fig.2, in the latter case, the actual excited FOV is indicated as a rectangular region. The hypo-intense regions in T2 (Fig.2a) indicate the likely position of the lesions, which is verified by the hyper-intense DCE perfusion signal in Fig.2b. Corresponding hyper-intense ASL perfusion regions as indicated by the pink arrows (1 and 2) are also observed in Fig.2c. The other two hyper-intense perfusion regions (blue arrow 4, 5) on FAIR ASL image are likely vascular regions. To verify the high perfusion nature of the lesion, a normal tissue region was selected as indicated by arrow 3. The unnormalized label and control signal level of repeated FAIR measurements for regions 1, 2, and 3 are shown in Fig.2d, clear differentiation of the perfusion signal level can be seen. Although TI of 1.2 s was shown, hyper-perfusion presented with all TI used. In this work, the feasibility of using reduced FOV ASL perfusion in prostate cancer is investigated. The hyper-perfusion region observed in rFOV ASL was consistent with that of conventional DCE, and shown as lesion in the T2 image. Not only rFOV excitation reduces susceptibility related artifacts while maintaining spatial resolution (In-plane resolution of 1.67mm was achieved that was comparable to that of 1.41mm in DCE), it is also able to avoid region with known artifacts such as motion or lipid. Sufficient SNR was obtained even with rFOV excitation due to sufficient averages, as seen by the distinct perfusion levels in lesion and healthy tissues. Although the use of TI was not specifically optimized, all the TIs used led to sufficient level of perfusion signal. It is hoped that the use of reduced FOV may help with the broad application of ASL based perfusion in body imaging.