In polygynous mammals, sex-specific patterns of body growth are linked to divergent selection pressures on male and female body size, resulting in sexual dimorphism (SD). For males, reproductive success is generally linked to body size, hence, males should prioritise early growth. For females, reproductive success is linked to resource availability, so they may adopt a more conservative growth tactic. Using longitudinal monitoring of known-age animals in two contrasting populations and an allometric approach to disentangle the relative contribution of structural size and physiological condition to SD, we addressed these issues in the weakly polygynous roe deer. Despite very different environmental conditions, we found remarkably similar patterns in the two populations in the mass-size allometric relationship at each life history stage, suggesting that relative allocation to structural size and physiological condition is highly constrained. SD in structural size (indexed by hind foot length) involved sex-specific growth trajectories governed by a single mass-size allometric relationship during the juvenile stage, such that males were both bigger and heavier than females. In contrast, SD in physiological condition (indexed by the allometric relationship between body mass and hind foot length, expressed as body mass for a given body size) developed markedly during the sub-adult stage in relation to sex differences in the timing of first reproduction. Among adults, males were heavier for a given size than females, suggesting that, relative to females, males express a capital breeder tactic, accumulating fat reserves to off set reproductive costs. By the senescent stage, SD in physiological condition had disappeared, with both sexes governed by a single allometric relationship, suggesting more rapid senescence in males than females. Individuals born into poor cohorts were generally lighter for a given size, indicating growth priority for skeletal size over physiological condition in both sexes. However, sex differences in cohort effects among sub-adults resulted in lower size-specific SD in poor cohorts, indicating that body condition of sub-adult females is buffered against environmental harshness. We conclude that sex-differences in reproductive tactics impose constraints on the ontogeny of SD in roe deer, leading to sex-specific trajectories in structural size and physiological condition.