Religious involvement is associated with mental health and well-being in non-military populations. This study examines the relationship between religiosity and PTSD symptoms, and the mediating effects of anxiety and depression in Veterans and Active Duty Military (V/ADM). This was a cross-sectional multi-site study involving 585 V/ADM recruited from across the USA. Inclusion criteria were having served in a combat theater and PTSD symptoms. Demographics, military characteristics, and social factors were assessed, along with measurement of religiosity, PTSD symptoms, depression, and anxiety. Bivariate and multivariate analyses examined the religiosity-PTSD relationship and the mediating effects of anxiety/depression on that relationship in the overall sample and stratified by race/ethnic group (White, Black, Hispanic). In bivariate analyses, the religiosity-PTSD relationship was not significant in the overall sample or in Whites. However, the relationship was significant in Blacks (r = - 0.16, p = 0.01) and in Hispanics (r = 0.30, p = 0.03), but in opposite directions. In the overall sample, religiosity was inversely related to anxiety (r = - 0.07, p = 0.07) and depression (r = - 0.21, p < 0.0001), especially in Blacks (r = - 0.21, p = 0.001, and r = - 0.34, p < 0.0001, respectively); however, in Hispanics, religiosity was positively related to anxiety (r = 0.32, p = 0.02) as it was to PTSD symptoms. When anxiety/depression was controlled for in multivariate analyses, the religiosity-PTSD relationship in the overall sample reversed from negative to positive, approaching statistical significance (B = 0.05, SE = 0.03, p = 0.079). In Blacks, the inverse association between religiosity and PTSD was explained by quality of relationships, whereas the positive relationship in Hispanics was explained by anxiety symptoms. In conclusion, religiosity was inversely related to PTSD symptoms in Blacks, positively related to PTSD in Hispanics, and unrelated to PTSD in the overall sample and in Whites. Anxiety/depression partially mediated the relationship in the overall sample and in Hispanics. Although longitudinal studies will be necessary to determine how these relationships come about, consideration should be given to spiritual/religious interventions that target anxiety/depression in V/ADM with PTSD.