Background Reliance on full breastfeeding alone for a longer time could have deleterious nutritional and health implications at later stages of children’s lives. About 47% of children are weaned at age =6 months and more than 50% of children in India under 4 years are stunted. We investigated the association between timing of weaning and stunting of children in India, using the data from National Family Health Survey, 1992–1993.

Methods Logistic analyses were employed on pooled data comprising one state each from six regions of India (N = 6285) with height status of children aged 2–4 years as the dependent variable. Timing of weaning was considered as the main control variable in the regression models.

Results Children weaned at age 6 months (odds ratio [OR] = 1.57) and after 6 months (OR = 1.88) were more likely to be stunted at later age compared with those weaned before 6 months (P < 0.001). Stunting appeared to be considerably lower for children weaned at age 3 months and showed an upward trend thereafter. The effect of age at weaning on stunting attenuated but persisted with statistical significance after controlling for important demographic, health, social and region variables. The likelihood of stunting was 77% for children weaned at age >6 months who had not received full immunization in the first year and had lived in poor conditions.

Conclusions Timing of weaning is significantly associated with stunting among children in India. The underlying causal associations between weaning behaviour and growth retardation need to be further examined by using longitudinal data