Puberty-driven increases in the secretion of testosterone may be a biological factor that protects males against the development of depression. Although all males produce testosterone, there are substantial between-person differences that could contribute to differential vulnerability to depression among pre-adolescent and adolescent boys, particularly after pubertal onset. Indeed, experimental animal and human data have shown that low testosterone increases risk for depressive-like symptoms in males, whereas higher levels of testosterone may be protective; however, prior studies have primarily investigated these effects in adulthood. This study investigated whether lower circulating levels of testosterone predict depressive symptoms in pre-adolescent and adolescent boys, and in particular, whether the testosterone-depression association becomes prominent with advancing pubertal maturation. Male twins (N = 213; ages 10-15 years) from the Michigan State University Twin Registry self-reported their depressive symptoms and pubertal status using the Children's Depression Inventory and the Pubertal Development Scale, respectively. Salivary testosterone was assayed using high-sensitivity enzyme immunoassays. Mixed Linear Models (MLMs), which could account for the non-independence of twin data, were used for analyses. As expected, lower testosterone concentrations were associated with higher depressive symptoms, and the magnitude of this effect increased with advancing pubertal status. In contrast, boys with higher levels of testosterone showed low levels of depressive symptomatology at all stages of pubertal maturation. Overall, these findings enhance understanding of within-sex variability in risk for depression in boys - average-to-high testosterone levels may underlie the general male resilience to depression after pubertal onset, whereas lower levels may increase vulnerability during/after puberty.