Literature review and cross-sectional study. Direct access to physical therapy necessitates greater responsibility to determine appropriateness of care by recognizing the potential for concomitant disease or systemic involvement. Recent research has identified excessive variability in the reporting of red flag symptoms. An initial step to improve the identification of red flag symptoms is the development of a standardized screening tool. To describe the development of a review-of-systems screening tool appropriate for use by orthopaedic physical therapists. First, a red flag symptom item bank was compiled from a systematic literature review to allow for further psychometric testing and development of a screening tool. Second, physical therapists in 11 outpatient clinics recruited patients presenting with primary complaints of neck, shoulder, low back, or knee pain. Patients completed the red flag symptom item bank and standard questionnaires for comorbidities, negative mood, quality of life, pain, and function. The development of the screening tool involved identifying and combining different 3-item sets that characterized the highest number of patients reporting at least 1 positive symptom response (operationally defined as "red flag symptom responder"). The literature search yielded 103 studies that met the inclusion criteria, and the final item bank consisted of 97 items representing 8 body systems. Four hundred thirty-one patients with primary complaints of neck (n = 93), shoulder (n = 108), low back (n = 119), or knee (n = 111) disorders contributed to the cross-sectional study. The number of red flag symptom responders was 393 of 431 (91.2%). These patients were older, more likely to be female, had lower income, and were more likely to report neck or back pain (all, P<.05). A 10-item review-of-systems screening tool correctly identified 372 of 393 (94.7%) responders, and a 23-item version identified all 393 (100%) responders. The review-of-systems screening tools and the complete 97-item bank had similar correlations with concurrent clinical measures, except for depressive symptoms. Concise red flag symptom identification appears to be feasible in outpatient orthopaedic physical therapy settings. Future research will determine how this review-of-systems screening tool needs refinement for different patient populations and whether it predicts clinical outcomes or the need for referral to other providers.