Early identification of autism spectrum disorder through cost-effective screening is crucial in low- and middle-income countries. The Child Behavior Checklist 1.5-5, using the Autism Spectrum Problems and Withdrawn Syndrome subscales, has potential as a level 1 autism spectrum disorder screening tool, though its construct validity in low- and middle-income countries remains underexplored. We aimed to validate the Child Behavior Checklist 1.5-5 for autism spectrum disorder screening in a representative sample of 1292 Brazilian children aged 3-5 years and 70 autism spectrum disorder children aged 1-5 years. Confirmatory factor analysis evaluated model fit indices and correlation strength between Child Behavior Checklist items and autism spectrum disorder diagnoses. Receiver operating characteristic curves assessed the optimal cutoff score. The Autism Spectrum Problems model demonstrated good fit and reliability (comparative fit index = 0.96, root mean square error of approximation = 0.037, [omega] = 0.869), as did the Withdrawn Syndrome model (comparative fit index = 0.974, root mean square error of approximation = 0.034, [omega] = 0.776), with one item per model showing low factor loadings. A cutoff score of 6 on the Autism Spectrum Problems yielded 82.5% sensitivity and 83.4% specificity, while a cutoff of 4 on the Withdrawn scale resulted in 87.9% sensitivity and 82.2% specificity. The Child Behavior Checklist 1.5-5 Autism Spectrum Problems and Withdrawn scales are reliable level 1 autism spectrum disorder screeners for Brazilian children, with good internal consistency and construct validity.