Background The Indian sex ratio has become highly male-biased in recent decades. This may be attributed to prenatal sex selection (PSS) and excess female infant mortality. However, the question of whether these factors are related has not been adequately studied. Here we examine whether increased use of PSS may offset excess female infant mortality, by reducing the number of ‘unwanted’ daughters being born.

Methods We analyse the National Family Health Survey (NHFS) data sets for India, which contain nationally representative samples of birth histories for women aged 15–49, interviewed in 1992–1993, 1998–1999 and 2005–2006. We test for missing female births at the second and third birth order, by analysis of the frequencies of sibling sex combinations, and examine the mortality differential between male and female infants, controlling for household wealth and sex(es) of older siblings.

Results PSS was used most in wealthier households at the second and third birth order, when the firstborn, or firstborn and second-born, siblings were female. Having preceding female siblings was a significant risk factor for female infant mortality, but was not correlated with household wealth.

Conclusions PSS and female infant mortality increase with the presence of older female siblings, yet we find no evidence that increasing use of PSS prevents female infant mortality, because PSS and the proportion of female infant mortality attributable to having older sisters increased over the study period. Increased pressure on higher birth order females caused by the trend towards smaller family sizes may explain this.