Background: Exclusive breastfeeding (EBF) is the single most cost-effective intervention to reduce infant mortality in developing countries. However, most studies concluded that a large proportion of infants are not exclusively breastfed. Objectives: to estimate duration of EBF using Kaplan-Meier analysis and to find out the independent factors associated with early cessation of EBF. Method: A cross-sectional descriptive study was carried out on 884 mother-infant pairs during vaccination at age of six months in Mansoura District, Egypt. Data collected are: sociodemographic characteristics of mothers, infants their families, antenatal, natal and postnatal data. Kaplan Meier estimator were used to produce valid estimates of EBF duration and the Cox's proportional hazards model was fitted to identify risks of early EBF discontinuation. Results: At six months 14.8% of infants are still exclusively breastfed. The median period of EBF was 4.0 months. The independent risk factors of EBF discontinuation are maternal work (Adjusted Hazard Ratio (AHR)=1.6, Confidence Interval (CI)=1.3-1.9), preterm births AHR=1.4, CI=1.1-1.1), urban residence (AHR=1.3, CI=1.1-1.5) and being female infant (AHR=1.2, CI=1.1-1.4). Conclusion: The median duration of EBF is below the recommended; and early discontinuation was associated with many factors that should be considered in designing breast feeding promotion programs.