Treatment of posttraumatic stress disorder (PTSD) is time and cost-intensive. New, readily implementable interventions are needed. Two parallel randomized clinical trials tested if cognitive/affective computerized training improves cognitive/affective functions and PTSD symptoms in acute (N = 80) and chronic PTSD (N = 84). Adults age 18–65 were recruited from an Israeli hospital emergency room (acute) or from across the United States (chronic). Individuals were randomized to an active intervention (acute N = 50, chronic N = 48) that adaptively trains cognition and an affective positivity bias, or a control intervention (acute N = 30, chronic N = 36) of engaging computer games. Participants, blind to assignment, completed exercises at home for 30 min/day over 30 days (acute) or 45 min/day over 45 days (chronic). Primary outcomes were computerized cognitive/affective function metrics. Secondary outcomes were Clinician-Administered PTSD Scale (CAPS) total scores. In chronic PTSD, the active arm demonstrated facilitated speed of fearful face identification (F = 20.96, q < 0.001; d = 1.21) and a trend towards improvement in total PTSD symptoms (F = 2.91, p = 0.09, d = 0.47), which was due to improvement in re-experiencing symptoms (F = 6.14, p = 0.015; d = 0.73). Better cognitive performance at baseline moderated the training effect and was associated with more favorable improvements on both metrics. Cognitive and affective training does not have widespread benefit on symptoms and cognitive/affective functions in PTSD. Future studies targeting re-experiencing a priori, stratifying on cognitive capacity, and with modified methods to infer on mechanisms and optimized training parameters may be warranted. ClinicalTrials.gov Identifiers: NCT01694316 &NCT02085512.