This paper estimates the causal effect of having young children aged 0–5 years on mothers’ labour force participation in rural India. To address the potential endogeneity in the fertility decision, I exploit Indian families’ preference for having sons. I leverage exogenous variation in the gender of older children aged 6 + years as an instrumental variable for having younger children aged 0–5 years in the family. IV estimates show that the mothers’ participation is significantly reduced by 9.9% due to the presence of young children aged 0–5 years in the household, with the negative effect mostly driven by mothers belonging to the highest income quartile; mothers with high education; and mothers residing in nuclear families. The findings highlight the need for investment in high-skilled jobs and formal childcare facilities to encourage mothers’ labour supply. Using the testable implications for the generalizability of LATE discussed in Angrist (The Economic Journal, 114: C52 C83, 2004), I show that the estimated causal effect is homogenous across compliers, always takers, and never takers and thus, generalizable to the whole population of interest.