It has been suggested that the personality of the individual influences appraisal processes that determine ratings of both social support and mental health, thus leading to an association between these variables that reflects only their shared self-report bias. To test whether ratings of social support made by someone other than the target individual would significantly predict outcomes, 115 pregnant adolescent girls and 115 adults who knew the adolescents well both completed ratings of the adolescent's social support. Results showed a moderate level of agreement between adolescent and informant support ratings. Both self-report and informant total social support ratings predicted pregnancy depression scores. Only informant social support ratings were significant predictors of postpartum depression scores. The ability of informants' ratings of social support to predict the adjustment of the adolescents to a major life stress was viewed as evidence that the commonly found link between social support and mental health reflects more than self-report biases.