Preterm birth is a leading cause of neurodevelopmental impairment in childhood^1,2, and the neural substrate for impairment includes diffuse white matter injury which is apparent from diffusion MRI (dMRI) in the newborn period. Chorioamnionitis (infection of the membranes that surround the fetus in utero) is associated with preterm birth in around 40% of cases³, and is an independent predictor of cerebral palsy in epidemiological studies⁴. However, there are uncertainties about its contribution to diffuse white matter injury associated with preterm birth, and its importance in relation to other injurious exposures experienced by preterm infants, including degree of prematurity and bronchopulmonary dysplasia (BPD)⁵. dMRI and tract-based spatial statistics (TBSS)^5,6 enable quantitative group-wise assessment of white matter microstructure in the newborn period. We combined histologic classification of chorioamnionitis in placental membranes with neonatal brain MRI, and used TBSS to test the hypothesis that chorioamnionitis is associated with altered white matter microstructure in preterm infants.Scanning: 26 preterm infants born with histopathological evidence of chorioamnionitis (defined in Redline et al⁷) at mean postmenstrual age (PMA) of 27 ± 6 weeks and 64 preterm infants born without chorioamnionitis at mean PMA of 29 ± 3 weeks were scanned on a MAGNETOM Verio 3 T clinical scanner (Siemens AG, Germany) at term equivalent age (PMA 40 ± 1 weeks) without sedation. All infants were scanned axially using a whole brain dMRI protocol consisting of 11 T2- and 64 diffusion-weighted (b = 750 s/mm²) single-shot spin echo EPI volumes with 2 mm isotropic voxels. Analysis: dMRI data were pre-processed using FSL tools (http://www.ndcn.ox.ac.uk/divisions/fmrib/). Fractional anisotropy (FA) maps constructed from this step were divided into groups born with or without chorioamnionitis. To identify a study-specific target image, every subject’s FA map was co-registered to every other subject’s and the target volume with the minimum displacement score from all other subjects in the group was deemed to be the most ‘typical’ image of the group (based on TBSS pipeline optimized for neonates⁵). Then, each subject’s FA map was aligned into the target space and upsampled to 1×1×1 mm³ voxel size. An average FA map and mean FA skeleton (thresholded at FA > 0.15) representing the center of all white matter tracts common to the group was created. Statistical comparison between the groups born with and without chorioamnionitis was performed with FSL’s Randomise using a general linear univariate model, with PMA at birth, PMA at scan and BDP listed as covariates. All images were subject to family-wise error correction for multiple comparisons following threshold-free cluster enhancement and are shown at p < 0.05.Preterm infants exposed to chorioamnionitis demonstrated decreased FA in the genu of the corpus callosum, cingulum cingulate gyri, centrum semiovale, inferior longitudinal fasciculi, anterior and posterior limbs of the internal capsule, external capsule and cerebellum (p < 0.05 corrected, Fig. 1). Skeleton wide FA was 5% lower in preterm infants with chorioamnionitis compared to infants born without this complication (Fig. 2).Chorioamnionitis is associated with diffuse white matter injury in preterm infants, and this is independent of known predictors for abnormal brain development after preterm birth. These data focus attention on the uterine environment for reducing brain injury associated with preterm birth.