The aim of this study was to develop and investigate the psychometric properties of a screening protocol for Norwegian students in upper secondary school. The protocol was designed to assess skills that are at stake in dyslexia. It was administered to 232 students. In the absence of a "gold standard," comparisons were made between students who reported normal literacy skills (n = 184) and literacy problems (n = 48). Significant group differences were found across all areas. Logistic regression and receiver operating characteristic curve analyses demonstrated good discriminatory power. The screening protocol met the standards for reliability and validity. It has the potential to be a useful tool for teachers to identify students at risk for dyslexia and who thus need to be referred to a full diagnostic investigation. Copyright © 2017 John Wiley & Sons, Ltd.