Objective: Loss of pleasure or interest in activities (i.e., anhedonia) is a risk factor for suicidality, treatment nonresponse, and relapse. Extant treatments that focus on reducing negative affect have limited effects upon positive affect (a core feature of anhedonia). We investigated whether a novel intervention aimed at increasing reward sensitivity was more efficacious for positive affect than a cognitive–behavior treatment aimed at reducing threat sensitivity, in individuals with clinically severe symptoms of depression or anxiety, and functional impairment. Method: The Treatment for Affective Dimensions trial was offered in a 2-site randomized study at outpatient treatment centers in Los Angeles and Dallas. Ninety-six patients were randomized to 15 weekly, individual sessions of Positive Affect Treatment (PAT) or Negative Affect Treatment (NAT). The primary outcome was improvement in positive affect (Positive and Negative Affect Schedule–Positive) from pretreatment to 6-month follow-up (6MFU). Secondary outcomes were improvements in negative affect (Positive and Negative Affect Schedule–Negative), suicidal ideation, and symptoms (Depression Anxiety Stress Scales). Results: PAT resulted in greater improvements in positive affect, p = .009, d = .52, and higher positive affect at 6MFU, p = .002, d = .67, than NAT. Participants in PAT also reported lower negative affect, p = .033, d = .52, and lower symptoms of depression, p = .035, d = .34, anxiety, p < .018, d = .30, and stress, p = .006, d = .43 at 6MFU. Finally, probability of suicidal ideation at 6MFU was lower in PAT than NAT (1.7% vs. 12.0%), p < .001. Conclusions: Compared to NAT, PAT demonstrated better outcomes (at 6MFU) on positive affect, depression, anxiety, stress, and suicidal ideation, for patients with symptomatic pretreatment levels of these outcomes.