Posttraumatic stress disorder (PTSD) impacts millions of Americans annually. Altered white matter microstructure may be a potential diagnostic biomarker for PTSD. White matter microstructural differences in persons with PTSD have been studied using machine learning, a method uniquely suited for biological datasets. This study examined the utility of white matter tracts in classifying persons with and without PTSD and predicting PTSD symptom cluster severity amongst trauma-exposed Black American women. Fractional anisotropy of 53 white matter tracts served as input features. Current PTSD presence was estimated using the Clinician-Administered PTSD Scale. Symptom cluster scores were calculated using the PTSD Symptom Scales. Only the random forest model demonstrated above-chance accuracy (58.88%) when classifying persons with and without PTSD. Regression models for symptom scores failed to show positive R-values. Results show a minimal signal for white matter microstructure and suggest a restricted set of white matter tracts is relevant to PTSD presence.