<h3>BACKGROUND:</h3> We evaluated the outcome of subjects with ARDS in relation to etiology and severity in a retrospective cohort study of the ARDS Network randomized controlled trials. The primary outcome was 28-d mortality. The secondary outcomes were 60-d mortality and ventilator- and ICU-free days. For severity of ARDS, subjects were stratified according to <b>P</b><sub>aO2</sub>/<b>F</b><sub>IO2</sub>. The etiology of ARDS was classified into sepsis, pneumonia, aspiration, trauma, and others. <h3>RESULTS:</h3> A total of 2,914 subjects were included in these trials. Outcomes were modeled with multivariable regressions adjusted for baseline covariates, age, sex, race, Acute Physiology and Chronic Health Evaluation III (APACHE III), vasopressor use, modified lung injury score, diabetes mellitus, cancer status, body mass index, pre-ICU location, ICU location, and study. There was no statistically significant difference in 28-d mortality in relation to ARDS severity. Subjects with trauma, compared with other etiologies of ARDS, had significantly lower mortality at 28 d (odds ratio [OR] = 0.47, 95% CI 0.26−0.83, <i>P</i> = .01). Sixty-day mortality was significantly lower for trauma subjects and those with severe ARDS group (OR = 0.5, 95% CI 0.3−0.85, <i>P</i> = .01 and OR = 0.71, 95% CI 0.52−0.98, <i>P</i> = .034, respectively). There were statistically significantly more ICU-free days and ventilator-free days for the aspiration group (OR = 1.09, 95% CI 1.02−1.17, <i>P</i> = .01 and OR = 1.09, 95% CI 1.02−1.16, <i>P</i> = .01, respectively). There was no statistically significant difference in ICU-free days or ventilator-free days in relation to severity of ARDS. <h3>CONCLUSIONS:</h3> Severity of ARDS based on <b>P</b><sub>aO2</sub>/<b>F</b><sub>IO2</sub> did not impact 28-d mortality, ventilator-free days, or ICU-free days. Among the etiologies of ARDS, trauma subjects had the lowest 28- and 60-d mortality, whereas subjects with aspiration had more ICU-free days and ventilator-free days.