Data collected during a prospective cohort study of infant feeding and health in rural Bilbeis, Egypt, were analysed to define prelacteal infant feeding practices, identify their determinants, and assess whether these practices were predictive of breastfeeding and supplementation patterns and diarrhoea incidence during infancy. Sixty per cent (89/149) of study infants were prelacteally fed sugar-water, teas, or both. Lack of milk in the mother's breast (74 per cent), and maternal exhaustion or illness following labour (29 per cent) were the two most commonly stated reasons for prelacteal feeding. After multivariate adjustment, significantly higher incidence of prelacteal feeding was associated with childbirth during the warmer months [odds ratio (OR): 2.4;95 per cent confidence interval (CI): 1.1–5.1], birth attendants with modern training (OR: 5.5; 95 per cent CI: 1.7–17.5), and labour lasting > 8 hours (OR: 2.3; 95 per cent CI: 0.1–4.9). Prelacteally fed infants were significantly less likely to be exclusively breastfed in age periods 0–3, 4–7, and 8–11 weeks. Diarrhoea incidence was higher among prelacteally fed infants in age periods 0–11, 12–23, and 36–47 weeks. Indiscriminate practice of prelacteal feeding and early supplementation of breastfeeding need to be discouraged.