The present study addressed cognitive biases in anxiety and depression using the emotional Stroop task, and explored both the affective space and autonomic underpinnings of these disorders. In previous studies, anxiety has been associated with both an attentional bias toward threat information and low cardiac vagal control, as reflected in heart rate variability (HRV) indices. Depression has been linked to a memory bias for negative information; however, findings of low HRV for depression are mixed. The high comorbidity of these disorders renders such findings as difficult to interpret. In the present study, it was hypothesized that the negative affect groups (anxious, depressed, comorbid anxious/depressed) would have lower vagally mediated HRV across tasks compared to the control group and that the anxiety and depression groups would show biases for group specific words on the Stroop task. Results for the Stroop tasks generally support previous findings of an attention bias in anxiety. The comorbid anxiety/depression group generally showed lower vagal control across tasks compared to the other groups, although comparisons between the "pure" anxiety and depression groups and the controls were not significant. It is suggested that this is because the comorbid group had higher depression and anxiety than either of the "pure" groups.