Objective To identify factors associated with nonalcoholic carbonated beverage (soft drink) consumption in children. Design Mail-in surveys collected by Dragonfly, a children's educational magazine distributed nationally to elementary and middle schools, were analyzed. The survey included questions about frequency of soft drink consumption and factors related to soft drink consumption. Subjects and participants The sample consisted of 560 children, 8 to 13 years old, who completed and mailed in the survey. There was an equal distribution of boys and girls (51% and 49%, respectively). Statistical analyses Frequency distributions were calculated and χ2 tests were conducted to determine whether soft drink consumption and related factors varied by sex and age. Multivariate logistic regression analyses were conducted to examine the association of soft drink consumption with each factor after adjustment for potential confounders. Results Preference for the taste of soft drinks was the strongest predictor in the analysis, with those who reported the strongest taste preference 4.50 times more likely (95% confidence interval=2.89-7.04) to consume soft drinks five or more times per week than those with a lower taste preference. Youth whose parents regularly drank soft drinks were 2.88 times more likely (95% confidence interval=1.76-4.72) to consume soft drinks five or more times per week compared with those whose parents did not regularly drink soft drinks. Conclusions Results suggest that several factors may be associated with soft drink intake in school-aged children, most notably taste preferences, soft drink consumption habits of parents and friends, soft drink availability in the home and school, and television viewing. Additional research is needed to verify these findings in a representative sample of children.