BACKGROUND: Expressive writing requires journaling stressor-related thoughts and feelings over four daily sessions of 15 min. Thirty years of research have popularized expressive writing as a brief intervention for fostering trauma-related resilience; however, its ability to surpass placebo remains unclear. This study aimed to determine the efficacy of expressive writing for improving post-traumatic stress symptoms in perinatal women who were living in the Houston area during major flooding caused by Hurricane Harvey. METHODS: A total of 1090 women were randomly allocated (1:1:1) to expressive writing, neutral writing or no writing. Interventions were internet-based. Online questionnaires were completed before randomization and at 2 months post-intervention. The primary outcome was post-traumatic stress symptoms, measured with the Impact of Event Scale-Revised; secondary outcomes were affective symptoms, measured with the 40-item Inventory of Depression and Anxiety Scales. Feelings throughout the intervention were reported daily using tailored questionnaires. RESULTS: In intention-to-treat analyses, no post-treatment between-group differences were found on the primary and secondary outcomes. Per-protocol analyses yielded similar results. A number of putative moderators were tested, but none interacted with expressive writing. Expressive writing produced greater feelings of anxiety and sadness during the intervention compared to neutral writing; further, overall experiences from the intervention mediated associations between expressive writing and greater post-traumatic stress at 2 months post-intervention. CONCLUSIONS: Among disaster-stricken perinatal women, expressive writing was ineffective in reducing levels of post-traumatic stress, and may have exacerbated these symptoms in some.