The child-therapist alliance is believed to be a critical ingredient of successful psychotherapy for youth depression. However, only a few studies have examined the association between the alliance and clinical outcomes in the treatment of youth depression. The present thesis examined the alliance-outcome association in two treatments for youth depression: cognitive-behavioral therapy (CBT) and usual clinical care (UC). Data were from an effectiveness trial conducted in six community clinics (see Weisz et al., 2009). Forty-one youth were randomly assigned to receive CBT or UC from community clinicians. The observed early alliance, alliance shifts, and self-reported alliance did not significantly predict child- or parent-reported depression outcome. However, the direction and strength of the alliance-outcome associations differed across alliance methodology (self- and observer-report) and condition (CBT and UC). Early child alliance did significantly predict treatment satisfaction. Implications and limitations of the results are discussed.