As a classical example of a sexually selected trait, the horns of male bovids offer a prime opportunity to identify predictors of the intensity of sexual selection. Here I use the comparative method to quantify sexual and natural selection pressures behind interspecific variation in horn length. I show that male horn length depends on factors proposed to affect the mean mate number per mating male, correlating positively with group size and negatively with male territoriality. This suggests that whereas group size increases the opportunity for sexual selection, territoriality reduces it because territorial males are unable to follow and monopolize female groups as effectively as males in nonterritorial species. Sexual body size dimorphism also correlates positively with group size and negatively with territoriality, corroborating these factors as predictors of the intensity of sexual selection on males. Female horn length was unaffected by the factors related to mating system, suggesting that this trait is mainly under natural selection. Using female horn length as a proxy for forces of natural selection revealed a negative effect on male horn length. Thus where natural selection favors female horns, possibly as effective weapons against predators, a similar selection pressure on males might prevent them from evolving too elaborate horns through sexual selection. There was no correlation found between horn length and latitude, thus providing no support for the hypothesis that horns have a thermoregulatory function.