The accuracy of parental reports of youth secondhand smoke exposure has received limited attention in the research literature. Such reports were compared to serum cotinine levels among participants of the National Health and Nutrition Examination Survey III who were aged 4-16 years. Likely smokers with serum cotinine values of 14 ng/ml or more and self-reported tobacco users were excluded from the analysis (n = 87), leaving 2,524 youth participants. One adult guardian, typically the parent, was asked to identify household smokers and estimate the number of cigarettes smoked in the home. Using cotinine levels of at least 0.2 ng/ml as the criterion, we found the sensitivity and specificity of any reported smokers in the home to be .65 and .92, respectively. Spearman correlations between cotinine levels and the number of smokers and the number of cigarettes consumed in the household were .67 and .68, respectively, and varied little across subgroups, including age, gender, and ethnicity of the child as well as household poverty status and educational attainment of the parent. Parental reports of household smoking alone fail to capture all youth secondhand smoke exposures, but they correlate well with cotinine levels when expressed as the number of household smokers or the number of cigarettes smoked in the household. Additional research is needed to determine whether reliance on parental reports of secondhand smoke exposure leads to bias in studies examining health outcomes in children and adolescents. Also, additional research is needed to better determine the level of secondhand smoke exposure that is biologically important in children and adolescents.