Multiple sclerosis (MS) is characterized by severe involvement of the spinal cord. At 1.5 T, spinal cord MS lesions have been described as rarely T1 hypointense [1], but the recent introduction of high-field scanners and optimized sequences improved the ability to detect T1 hypointensites in the upper cord [2,3]. While preliminary studies at 3.0 T suggested that both upper cord lesion load and cord atrophy contributed to disability [3], a full characterization of spinal cord lesions on high-resolution T1-weighted scans along the entire cervical cord segment is still missing. Moreover, the possible influence of T1 hypointensities on cord atrophy estimation, cord tissue loss and disability has never been explored. Against this background, aims of our study were: 1) to characterize the spatial distribution of cervical cord T1 lesions in a relatively large cohort of MS patients; 2) to assess whether there was any influence of cord T1 lesions on active surface (AS) [4] cord outline estimates; and 3) to analyze the association between cord T1 lesions, atrophy and disability. Sagittal 3D T1-weighted scans of the cervical cord were acquired at 3.0 T from 133 patients with MS (63 relapsing remitting [RR], 30 secondary progressive [SP], 20 primary progressive [PP] and 20 benign [B] MS) and 47 controls. T1-hypointense lesions were identified by an experienced neurologist on the sagittal 3D T1 images. Lesions were counted and binary lesion masks were produced. Then, the AS method was applied to calculate cross-sectional area (CSA) between C1 and C7. The quality of cord outline estimates was carefully reviewed. Unfolded cervical cord images were created by reformatting the input images perpendicular to the estimated cord centre line. Straightened images were coregistered to the average cord image of healthy controls, serving as a cervical cord template [5]. The same transformation was applied to the cervical cord masks and T1 lesion masks. Between-group comparisons of T1 lesions and cord atrophy were performed with ANOVA models adjusted for age, using SPSS and SPM12 (global and regional analysis, respectively). Multiple regression models were used to assess correlations between T1 cord lesion number and extent, cord atrophy and patient clinical disability.T1 hypointense lesions were detected in 114 (85%) of MS patients (53 RRMS, 28 SPMS, 16 PPMS and 17 BMS). The median number of T1 lesions was 3 (range=0-11 lesions) and the average cord T1 lesion volume was 0.3 ml (SD=0.3 ml). T1 lesion number and volume were not significantly different between phenotypes (p=0.08 and 0.4, respectively). The average T1 lesion probability maps in MS patients (as a whole) and in the different phenotypes are shown in Figure 1. T1 lesions were more frequently occurring in SPMS vs RRMS, and in PPMS vs RRMS and SPMS patients, especially in anterior and lateral upper cord segments (Figure 1). Cord outlines were correctly estimated by the AS method both in patients without and with T1 lesions (Figure 2). Significant cord tissue loss was found in MS patients vs controls, and 1) in RRMS and PPMS vs controls (p<0.001); 2) in RRMS vs BMS (p=0.01); and 3) in SPMS vs RRMS patients (p=0.002). Cord atrophy did not differ between PPMS and SPMS patients (Figure 3). Whole-cord CSA was not correlated with cord T1 lesion number/volume (p=0.8 and 0.3, respectively). The regional distribution of cord atrophy was modestly correlated with T1 lesion number/volume (Figure 4). Conversely, there was a strong correlation between cord atrophy and disability, both at a global (r=-0.51, p<0.001) and regional analysis (Figure 4). The association between cord T1 lesions and EDSS was also significant, both at a global (r=0.27, p=0.002) and regional level (Figure 4).T1 hypointense cervical cord lesions were detected in a large proportion (85%) of MS patients, with a slightly higher frequency in the progressive than in the relapsing forms of the disease. This widespread presence of T1 lesions did not influence cord area estimates produced by the AS method, which were reliable in all study subjects. This might be explained by the continuity constraints imposed by the AS method on cord surface smoothness along the cord axis direction, and by the good contrast between cord tissue and surrounding CSF [1]. The association between cord T1 lesions and cord atrophy was modest. However, both cord T1 lesions and atrophy were significant and independent contributors to patient physical disability.