Abstract While senescence is a common occurrence in wild populations, not all traits decline with age simultaneously and some do not show any senescence. A lack of senescence in secondary sexual traits is thought to be due to their importance for reproductive success. However, if reproductive success senesces, why would secondary sexual traits apparently not senesce? Here, we explored this question in a wild population of red deer (Cervus elaphus) using antler form (number of points), a secondary sexual trait which shows little senescence, despite the occurrence of reproductive senescence. In line with expectations for traits that senesce, genetic variance in antler form increased with age and selection weakened with age. Therefore, there was no indication that a stronger selection on individuals that survived to older ages was countering the dilution of selection due to fewer individuals being alive. Furthermore, the effect of selective disappearance masking a slight decline in antler form in the oldest years was small. Interestingly, although genetic variance and positive selection of antler form were found, there was no evidence of a response to selection, supporting a genetic decoupling of antler senescence and reproductive senescence. Finally, a positive genetic covariance in antler form among age classes provides a possible explanation for the lack of senescence. These findings suggest that the antler form is under a genetic constraint that prevents it from senescing, providing an interesting evolutionary explanation for negligible senescence in a secondary sexual trait, and consequently, the existence of asynchrony in senescence among traits within populations.