Objective. To determine the relationship between pacifier use at 1 month of age to the duration of breast-feeding to 6 months of age. Design. Longitudinal study of infants from birth to 6 months of age. Population. Six hundred five rooming-in infants born at the largest hospital in Guarujá, São Paulo, Brazil, during January and February 1993. Main outcome measures. Prevalence of breast-feeding (exclusive, predominant, and complementary) at 1, 4, and 6 months of age. Results. Relative risk for weaning between 1 and 6 months of age was 3.84 (95% confidence interval 2.65-5.50) for pacifier users at 1 month of age, compared with nonusers. When an adjustment was made for possible confounding variables through Cox regression analysis, the relative risk dropped to 2.87 (95% confidence interval 1.97-419). Conclusion. Pacifier use is highly correlated with early weaning, even after controlling for possible confounders. Until it is determined if pacifier use is causally related to weaning or is a marker for other undetermined causes, pacifier use probably should not be recommended for breast-fed infants.