AIM: Diagnosing individuals at ultra high risk (UHR) for psychosis can improve early access to treatment, and a two-stage model utilizing self-report screening followed by a clinical interview can be accurate and efficient. However, it is currently unclear which screening cut-offs to adopt with different populations. METHODS: A systematic review of diagnostic accuracy studies evaluating the Prodromal Questionnaire (PQ) as a preliminary screener for UHR and psychosis was conducted to examine screening effectiveness in different contexts. MedLine, PsycInfo, SCOPUS and ProQuest Dissertations and Abstracts databases were electronically searched, along with a review screen and citation search of key papers. Findings were summarized in a narrative synthesis. RESULTS: In total, 14 diagnostic accuracy studies and 45 studies using the PQ as a screening tool for UHR and psychosis were included. In all settings, the 3 different versions of the PQ were all found to accurately identify UHR and full psychosis. Higher cut-off points were required in non-help-seeking samples, relative to general help-seeking populations, which in turn were higher than those needed in samples highly enriched with UHR participants. CONCLUSION: The findings support the use of the PQ as a preliminary screening tool for UHR in different settings; however, higher thresholds in lower UHR-prevalence populations are necessary to minimize false positives. Including the distress criteria, rather than just number of symptoms, may improve screening effectiveness. Different thresholds may be appropriate in different contexts depending on the importance of sensitivity vs specificity. Protocol registration: CRD42016033004.