Trauma cue-elicited activation of automatic cannabis-related cognitive biases are theorized to contribute to comorbid posttraumatic stress disorder and cannabis use disorder. This phenomenon can be studied experimentally by combining the trauma cue reactivity paradigm (CRP) with cannabis-related cognitive processing tasks. In this study, we used a computerized cannabis approach-avoidance task (AAT) to assess automatic cannabis (vs. neutral) approach bias following personalized trauma (vs. neutral) CRP exposure. We hypothesized that selective cannabis (vs. neutral) approach biases on the AAT would be larger among participants with higher PTSD symptom severity, particularly following trauma (vs. neutral) cue exposure. We used a within-subjects experimental design with a continuous between-subjects moderator (PTSD symptom severity). Participants were exposed to both a trauma and neutral CRP in random order, completing a cannabis AAT (cannabis vs. neutral stimuli) following each cue exposure. Current cannabis users with histories of psychological trauma (n = 50; 34% male; mean age = 37.8 years) described their most traumatic lifetime event, and a similarly-detailed neutral event, according to an established interview protocol that served as the CRP. As hypothesized, an AAT stimulus type x PTSD symptom severity interaction emerged (p = .042) with approach bias greater to cannabis than neutral stimuli for participants with higher (p = .006), but not lower (p = .36), PTSD symptom severity. Contrasting expectations, the stimulus type x PTSD symptoms effect was not intensified by trauma cue exposure (p = .19). Selective cannabis approach bias may be chronically activated in cannabis users with higher PTSD symptom severity and may serve as an automatic cognitive mechanism to help explain PTSD-CUD co-morbidity.