Introduction <p>Low health literacy (HL) can impair decision-making in emergencies. Extending our preceding adult study, this preregistered observational study examined whether parental HL affects accuracy of severity assessments in the pediatric emergency department (PED) and whether the validity of children’s self-reports is age dependent.</p> Methods <p>This single-center prospective cross-sectional study enrolled 182 pediatric patients and parents presenting to a German university hospital PED. Subjective and independent emergency condition severity assessments were obtained from patients, parents, and medical staff. Parental HL was measured using the 16-item European Health Literacy Survey questionnaire (HLS-EU-Q16). Reference standard severity assessment was retrospectively collected by specialists (retrospective chart-based chase assessment). Analyses compared assessment accuracy across three parental HL level groups (adequate, problematic, inadequate). Assessments were adjusted for covariates except for regression analyses, which we performed using unadjusted values.</p> Results <p>Parental HL showed no measurable effect on assessment accuracy, discrepancy or concordance with medical team’s evaluations. Children’s self-reported anxiety emerged as a stronger predictor of actual clinical severity than HL measurements, with this relationship becoming significant at 8 years of age. Children’s severity assessment (OR = 1.38 [1.20, 1.57], <i>p</i> &lt; .001) and less distinctly anxiety (OR = 1.15, 95% CI [1.01, 1.31], <i>p</i> = .037) predicted severe outcome, while the interaction between age and anxiety (B = 0.071, SE = 0.031, <i>p</i> = .022) showed that anxiety becomes a reliable indicator at around early school age. Concordance with the team’s assessment likewise emerged at school age (ρ ≈ 0.30, <i>p</i> &lt; .05). Direct children’s self-assessment provided clinically meaningful information across all HL groups. Nursing assessments remained the strongest predictor of severe outcomes (OR = 1.95, 95% CI [1.37, 2.67], <i>p</i> &lt; .001), followed by physicians (OR = 1.62, 95% CI [1.20, 2.19], <i>p</i> &lt; .001). Parents were in our study not able to predict severe outcome (<i>p</i> = .22). Parental underestimation relative to specialists was consistently linked to severe outcomes (26–33%), regardless of HL level.</p> Conclusion <p>Contrary to common assumptions, parental HL neither impaired nor improved recognition of pediatric emergency severity, and HL did not systematically shift discrepancy patterns. Instead, developmental maturity and professional observation are relevant for severity assessment. From early school age (7–8 years), children’s anxiety appears to provide an increasingly valid, dose-responsive signal of severity. Nurses’ assessments provided the highest predictive accuracy, underscoring their central role in triage. Age-appropriate use of pediatric patients’ self-assessments could further support risk stratification in the PED.</p> Clinical trials registration <p>This observational study was preregistered in the German Clinical Trials Register (DRKS): DRKS00034108, Registration Date: 30th January 2025.</p>