ObjectiveTo evaluate the validity of the triage system CLARIPED in a pediatric population in the city of São Paulo, Brazil. MethodsProspective, observational study in a secondary-level pediatric emergency service from Sep-2018 to Ago-2019. A convenience sample of all patients aged 0–18 years triaged by the computerized CLARIPED system was selected. Associations between urgency levels and patient outcomes were analyzed to assess construct validity. Sensitivity, specificity, and positive and negative predictive values (PPV and NPV) to identify the most urgent patients were estimated, as well as under-triage and over-triage rates. ResultsThe distribution of 24,338 visits was: RED 0.02%, ORANGE 0.9%, YELLOW 23.5%, GREEN 47.9%, and BLUE 27.7% (highest to the lowest level of urgency). The frequency of the following outcomes increased with increasing urgency: hospital admission (0.0%, 0.02%, 0.1%, 7.1% and 20%); stay in ED observation room (1.9%, 2,4%, 4.8%, 24.1%, 60%); use of ≥ 2 diagnostic/therapeutic resources (2.3%, 3.0%, 5.9%, 28.8%, 40%); ED length of stay (12, 12, 15, 99.5, 362 min). The most urgent patients (RED, ORANGE, and YELLOW) exhibited higher chances of using ≥ 2 resources (OR 2.55; 95%CI: 2.23–2.92) or of being hospitalized (OR 23.9; 95%CI: 7.17–79.62), compared to the least urgent (GREEN and BLUE). The sensitivity to identify urgency was 0.88 (95%CI: 0.70–0.98); specificity, 0.76 (95%CI: 0.75–0.76); NPV, 0.99 (95%CI: 0.99–1.00); overtriage rate, 23.0%, and undertriage, 11.5%. ConclusionThis study corroborates the validity and safety of CLARIPED, demonstrating significant correlations with clinical outcomes, good sensitivity, and low undertriage rate in a secondary-level Brazilian pediatric emergency service.