The objective of this meta-analysis was to determine with cross-sectional and cohort trails whether the use of pacifier increases the risk of early weaning from exclusive breastfeeding before six months of age or cessation of breastfeeding from any breastfeeding before 24 months of age. Additionally, the effect of the age for starting pacifier use on breastfeeding duration was analyzed in the cohort trails. The Medline database was searched (1980 to 2006) with "breastfeed, breastfeeding, or breast feed" and "pacifier, dummy, or soother" as individual keywords. Only human studies published in English were included. Unpublished data were not sought. Twelve trials with weaning from exclusive breastfeeding and 19 trials with cessation of any breastfeeding were included in the meta-analysis. The meta-analysis was performed with Stata 6.0 statistical package. Summary risk ratio for early weaning before six months of age in exclusive breastfeeding trails was 2.016 (95% CI: 1.619-2.511) for pacifier users compared with nonusers in studies with univariate analysis and 1.792 (95% CI: 1.452-2.212) in studies with multivariate analysis. Similarly, pacifier usage compared with nonusers reduced the duration of any breastfeeding in both univariate (2.760, 95% CI: 2.083-3.657) and multivariate trials (1.952, 95% CI: 1.662-2.293). The use of pacifiers was associated with shortened duration of exclusive and of any breastfeeding. Given the increase in the benefits with duration of breastfeeding, parents should be informed of the link between pacifier use and shortened breastfeeding duration in order to help them make informed decisions about their children's care.