In the context of high neonatal mortality rate (NMR) in developing country settings, a promising strategy for enhancing newborn health is promotion of preventive newborn care practices. We measured the effect of a behaviour-change intervention on perceived neonatal illnesses in rural Uttar Pradesh, India. The study was nested in a cluster-randomized controlled trial of the impact of a package of essential newborn care on NMR. We prospectively enrolled 802 mothers and administered a questionnaire on perceived neonatal morbidities. Regression analysis showed that newborns in the intervention clusters had significantly lower risk of perceived diarrhoea [adjusted relative risk (aRR) 0.67, 95% confidence interval (CI) 0.49-0.90] and skin-related complications [aRR 0.67, 95% CI 0.45-1.00] compared to newborns in the comparison area. Assuming incidence of perceived illnesses is a proxy for actual morbidity rates, we conclude that promotion of preventive care practices through behaviour-change interventions was effective in reducing neonatal morbidities.