Background
This paper examines seemingly contradictory evidence from extant research that son preference is high, but male disadvantage in mortality is increasing in Nepal. To do so, we documented the timing, geographic patterning, and extent of gendered patterns in mortality and feeding practices for children under-five.

Methods
We applied pooled multilevel regression models and survival analysis to five rounds of data from Nepal’s nationally representative Demographic and Health Surveys (1996–2016). We controlled for potential sociodemographic confounders, including child, maternal, household, and regional correlates, and disaggregated findings by birth order and sibling gender.

Results
We found evidence of regional variation in mortality, with girls in wealthy urban areas faring the worst in terms of mortality rates. Girls’ comparative mortality advantage compared to boys in the neonatal period masks their mortality disadvantage in later periods. Mortality has fallen at a faster rate for boys than girls in most cases, leading to widening of gender inequalities. We also found evidence of female disadvantage in breastfeeding duration, which was linked to higher mortality risks, but no gender disparities in the consumption of other food items. Sibling gender and birth order also mattered for breastfeeding duration: Young girls with older sisters but with no brothers were most disadvantaged.

Conclusion
While we did not find evidence of postnatal discrimination in access to solid and semi-solid foods, girls in Nepal face a disadvantage in breastfeeding duration. Girls with older sisters but no older brothers facing the greatest disadvantage, with risks being particularly concentrated for girls aged 1–4 years. This disadvantage is linked to an increased risk of mortality. To address this, community-based health programs could be expanded to continue targeted healthcare for children beyond 12 months of age, with particular focus on nutrition monitoring and health service provision for girls.