While seasonal trends in testosterone levels are known from cross-cohort studies, data on testosterone inter-annual individual repeatability in wild birds are rare. Also, our understanding of hormonal age-dependent changes in testosterone levels is limited. We assessed plasma testosterone levels in 105 samples originating from 49 repeatedly captured free-living great tits (Parus major) sampled during the nesting to investigate their relative long-term repeatability and within-individual changes. Furthermore, we examined the inter-annual repeatability of condition-related traits (carotenoid-and melanin-based plumage ornamentation, ptilochronological feather growth rate, body mass, and haematological heterophil/lymphocyte ratio) and their relationships to testosterone levels. We show that testosterone levels are inter-annually repeatable in females, with a non-significant pattern in males, both in absolute values and individual ranks (indicating the maintenance of relative status in a population). In males, we found a quadratic dependence of testosterone levels on age, with a peak in midlife. In contrast, female testosterone levels showed no age-dependent trends. The inter-annual repeatability of condition-related traits ranged from zero to moderate and was mostly unrelated to plasma testosterone concentrations. However, males with elevated testosterone had significantly higher carotenoid-pigmented yellow plumage brightness, a trait presumably involved in mating. Showing inter-annual repeatability in testosterone levels, this research opens the way to further understanding the causes of variation in condition-related traits. Based on a longitudinal dataset, this study demonstrates that male plasma testosterone undergoes age-related changes that may regulate resource allocation. Our results thus suggest that, unlike females, male birds undergo hormonal senescence similar to mammals.