In the Children of the National Longitudinal Survey of Youth (NLSY-C), mothers rated their 12–23month old toddlers on 11 temperament items. Three sets of items (three items per set) – one clustered around fearfulness, one clustered around positive affect, and one clustered around difficultness – were subjected to taxometric analysis using mean above minus below a cut (MAMBAC), maximum covariance (MAXCOV), and latent-mode factor analysis (L-Mode). The results for all three sets of items showed consistent support for dimensional latent structure. When all nine items were simultaneously analyzed with finite mixture modeling the results were inconsistent with a categorical solution. The results of this study indicate that individual differences in childhood temperament – as measured by maternal ratings of children 12–23months of age – are quantitative (difference in degree) rather than qualitative (difference of kind). The implications of these results for understanding and assessing childhood temperament are discussed.