Weaponry in ungulates may be costly to grow and maintain, and different selective pressures in males and females may lead to sex‐biased natural survival. Sexual differences in the relationship between weapon growth and survival may increase under anthropogenic selection through culling, for example because of trophy hunting. Selection on weaponry growth under different scenarios has been largely investigated in males of highly dimorphic ungulates, for which survival costs (either natural or hunting related) are thought to be greatest. Little is known, however, about the survival costs of weaponry in males and females of weakly dimorphic species. We collected information on horn length and age at death/shooting of 407 chamois Rupicapra rupicapra in a protected population and in two hunted populations with different hunting regimes, to explore sexual differences in the selection on early horn growth under contrasting selective pressures. We also investigated the variation of horn growth and body mass in yearling males (n = 688) and females (n = 539) culled in one of the hunted populations over 14 years. The relationship between horn growth and survival showed remarkable sexual differences under different evolutionary scenarios. Within the protected population, under natural selection, we found no significant trade‐off in either males or females. Under anthropogenic pressure, selection on early horn growth of culled individuals showed diametrically opposed sex‐biased patterns, depending on the culling regime and hunters’ preferences. Despite the selective bias between males and females in one of the hunted populations, we did not detect significant sex‐specific differences in the long‐term pattern of early growth. The relationship between early horn growth and natural survival in either sex might suggest stabilizing selection on horn size in chamois. Selection through culling can be strongly sex‐biased also in weakly dimorphic species, depending on hunters’ preferences and hunting regulations, and long‐term data are needed to reveal potential undesirable evolutionary consequences.