Psychotic-like experiences (PLEs) are perceptual and thought disturbances that, although common among children, increase risk for future psychopathology, particularly if persistent. Clinical interviews are too time-consuming and costly to administer at a population level, but the criterion validity of a brief questionnaire for screening community samples of children as young as 9 years for PLEs has not been established. This study aimed to test the criterion (concurrent and predictive) validity of the Psychotic-Like Experiences Questionnaire for Children (PLEQ-C). The PLEQ-C (9-item self- and 10-item parent-report versions) was administered to 139 children aged 9–12 years and their caregivers recruited from Greater London, UK. Children additionally completed a diagnostic interview assessing hallucinations and delusions and three further PLEQ-C assessments at approximately 24-month intervals. Concordance of child- and caregiver-reports of PLEs on questionnaire (PLE-Q) was low. Self-reports of any PLE-Q demonstrated good sensitivity (73.3%), specificity (78.5%), positive and negative predictive values (PPV: 72.1%; NPV: 79.5%) for any PLE determined by interview (PLE-I), whereas caregiver-reports of any PLE-Q performed poorly (sensitivity 51.7%, specificity 78.5%, PPV 64.6%, NPV 68.1%). Multinomial regression analyses indicated that children reporting any PLE-Q at screening were at significantly increased risk of reporting PLEs on multiple assessments during adolescence relative to no PLEs, closely replicating the pattern and magnitude of effects (large-to-very large) obtained for children with any PLE-I. The PLEQ-C offers a valid, brief, feasible, and cost-effective means of community screening to identify children who present with PLEs and could be assessed with clinical interview.