ABSTRACT The Trivers‐Willard (1973) model suggests maternal control of offspring sex, in utero or by the end of parental investment, may be an adaptive advantage in some species. We tested for differential sex allocation using 11,408 known‐sex fetal elk (Cervus elaphus) from biological collections and hunter harvest returns from 2 southwestern Montana, USA, elk populations (1961–2007). We included maternal and environmental condition covariates measured pre‐ and postconception and throughout pregnancy. Results suggested that adult female elk in southwest Montana did not differentially invest in male offspring when conditions were beneficial. We found evidence that, when the Northern Yellowstone elk herd was at low density, beneficial spring (May‐Jun) growing conditions, as indexed by a local precipitation measure and a regional drought indicator, correlated with production of more female fetuses (1 SD increase in precipitation and 1 SD decrease in drought resulted in 6% and 5% more F fetuses, respectively). In the same herd, we found evidence that improved maternal condition, as indexed by kidney fat mass and heart fat mass, also correlated with production of more female fetuses (1 SD increase in kidney fat mass and heart fat mass resulted in 8% more F fetuses). When the same elk herd reached higher densities under different ecological conditions, no covariate was associated with a deviation in the 50:50 female‐to‐male sex ratio. Similarly, there was no association between covariates and fetal sex ratios in a nearby elk herd at high population density. In modeling, wildlife managers should consider factors that could alter sex ratios at birth, and also how biased sex ratios postpartum could affect population models.