With advances in both software (optimal coil and sequences) and hardware high-resolution MR imaging for brachial plexus will become more common. Although 3D fast spin-echo (FSE) T2-weighted image and IDEAL water image result in good fat suppression and delineation of peripheral nerves owing to increased dynamic range of contrast, the imaging is frequently limited by suboptimal signal-to-noise ratio and conspicuous vascular signal, which limits the selective evaluation of smaller peripheral nerves within their neurovascular bundles. Most of the applications of double inversion recovery (DIR) technique have been especially for detection of multiple sclerosis lesions of the cerebral cortex [1] and carotid vessel wall imaging [2]. In this study, we propose a new scheme to acquire high-resolution and volumetric neurographic images using 3D fat suppression T2-weighted Cube combined with DIR (3D FS DIR Cube). It employs both inversion recovery technique and frequency-selective fat saturation to obtain more robust homogeneous fat suppression. Moreover, Cube, variable reforcus flip angle FSE can be generated with excellent vascular suppression because of sufficient blood flow masking effect [3].The aim was to evaluate the feasibility of brachial plexus assessment using 3D FS DIR Cube and to compare it with 3D FS T2-weighted Cube (3D FS T2 Cube) and 2D IDEAL water.Experiment – 15 normal volunteers (10 men and 5 women; 23-51 years, mean 35.1) were imaged on a 3T MRI scanner (Discovery 750w, GE Healthcare, Waukesha, WI, USA) using a GEM NV coil. 3D FS DIR Cube, 3D FS T2 Cube, and 2D IDEAL water were acquired. The scan parameters for the 3D FS DIR Cube were as follows: TR/TE 5100/60, IR1/IR2 2200/220, +50 kHz BW, 2.0mm slice thickness, 320 × 224 matrix, ETL 120, 1 NEX, 34 cm FOV, and scan time 4 mm 18 seconds. Scan parameters for 3D FS T2 Cube were as follows: TR/TE 3200/maximum, 50 kHz BW, 2 mm slice thickness, 320 × 224 matrix, ETL 96, 1 NEX, 34 cm FOV, and scan time 3 minutes 58 seconds. And scan parameters for 2D IDEAL were as follows: TR/TE 4901/68, +83.33 kHz BW, 3 mm slice thickness, 320 × 224 matrix, ETL 13, 2 NEX, 34 cm FOV, and scan time 4 minutes 10 seconds. Image Evaluation - Two musculoskeletal radiologists evaluated all images in consensus. The images were assessed for nerve-to-fat signal ratio (nerve signal/fat signal), nerve-to-muscle signal ratio (nerve signal/muscle signal), nerve-to-cerebrospinal fluid (CSF) signal ratio (nerve signal/CSF signal), nerve-to-disc signal ratio (nerve signal/intervertebral disc signal), and nerve-to-artery signal ratio (nerve signal/subclavian artery signal). Nerves were measured to calculate the means of four different regions (bilateral C6 and C7 nerve roots). Intervertebral disc was measured at the level of C6/7. The conspicuity of brachial plexus branch, homogeneity of fat suppression, vessels signal suppression, and quality of maximum intensity projection (MIP) image were qualitatively assessed. The Kruskal-Wallis test and the Mann-Whitney U test with Bonferroni correction were used to compare the mean values obtained from each sequence. A value of p < 0.05 was considered significant.3D FS DIR Cube had the highest nerve-to-fat signal ratio and nerve-to-CSF signal ratio that was significantly higher than 3D FS T2 Cube and 2D IDEAL water. 3D FS DIR Cube had also the highest nerve-to-artery signal ratio that was significantly higher than 2D IDEAL water (Fig. 1). In qualitative assessment, the conspicuity of brachial plexus branch, homogeneity of fat suppression, vessels signal suppression, and quality of MIP was best on 3D FS DIR Cube (Fig. 2, 3).The 3D FS DIR Cube can provide excellent contrast between brachial plexus and surrounding tissues and increase the conspicuity of the nerves delineation with uniformed fat suppression and vessels signal suppression compared to 3D FS T2 Cube and 2D IDEAL water.