Our study has considerable strengths, including the use of a multistage probability sampling design for subject recruitment and a sample size that is larger than that of any previous study of health literacy and childhood asthma. We also recognize several limitations to our findings. First, our study was cross-sectional, and we are thus unable to assess causal or temporal relationships. For example, the observed association between reported use of ICS and ED/urgent care visits or hospitalizations is likely due to “reverse causation” (eg, if only children with more severe asthma were prescribed ICS). Second, we lack data on adherence to prescribed ICS. Third, low parental numeracy is tightly linked to poverty, and there could thus be residual confounding by socioeconomic status, asthma severity, or other unmeasured factors (eg, housing quality, health beliefs, or cultural patterns). However, confounding by socioeconomic status or disease severity is an unlikely explanation for our findings, because we obtained similar results in multivariate models adjusting for various indicators of socioeconomic status (household income, type of health insurance, or parental education) or asthma severity/control (FEV1/FVC, one or more positive allergen-specific IgE, or use of leukotriene inhibitors). Fourth, we did not measure child’s numeracy, which may be relevant to older children taking most of the responsibility for their asthma care. Fifth, although the ANQ is a validated test, the modified version used for our analyses was not previously validated. Finally, our results may not be generalizable to Puerto Rican children living in the mainland United States (as individuals who emigrate may differ from those who do not) or children in other ethnic groups. However, Puerto Rican children living in the mainland United States are more likely to come from families with a lower socioeconomic status3 and, therefore, are more likely to be exposed to low parental numeracy.