Background Respiratory therapists (RTs) need continuous professional development to stay competent in evolving clinical areas. However, there is no validated, professional-specific tool for systematically evaluating their learning needs in respiratory therapy. This study aimed to create and preliminarily validate the Respiratory Therapy Learning Needs Instrument (RT-LNI), a self-assessment based on key competency frameworks.Methods A two-phase cross-sectional study was carried out at three Dubai Health tertiary care centers. In the first phase, six expert RTs assessed 83 items across 13 competency domains using the content validity index (CVI) methodology. During the second phase, 58 licensed RTs were invited and 50 completed the refined 75-item instrument (response rate of 72.5%), along with a user evaluation survey completed by 45 participants. Psychometric evaluation included Cronbach's alpha and domain-level exploratory factor analysis (principal axis factoring with varimax rotation; promax sensitivity analysis). The number of factors was determined by parallel analysis. Subgroup comparisons employed Mann-Whitney U tests with Bonferroni correction. The reporting adhered to the STROBE checklist.Results Content validation demonstrated high validity, with an S-CVI/Ave of 0.90 (95% CI 0.86–0.94). After removing eight items, the instrument consisted of 75 items. Internal consistency ranged from acceptable to excellent (Cronbach's alpha 0.70–0.95; median 0.91). Domain-level factor analysis mostly indicated unidimensional structures (KMO 0.64–0.88; median primary factor loading 0.79), except in Emergency and Critical Care, which showed a two-factor structure—acute resuscitation versus prolonged management. Self-assessed competence was highest in core technical skills (mean 3.7–3.9/5, corresponding to Dreyfus "Proficient" stage) and lowest in evidence-based practice (2.4/5), sleep medicine (2.6/5), and pulmonary rehabilitation (2.7/5), at the "Advanced Beginner" stage. No significant subgroup differences emerged after Bonferroni correction, although the study was underpowered for small-to-medium effects. User feedback indicated strong acceptability, with a mean score of 4.26/5 (85.3%).Conclusions The RT-LNI demonstrated strong content validity, acceptable psychometric properties, and high user acceptability in this preliminary validation. As the first published, psychometrically evaluated, profession-specific learning needs assessment instrument for respiratory therapy, the tool fills a gap in the CPD infrastructure. The identified learning needs patterns have direct implications for targeted workforce development. Multi-site confirmatory validation with larger samples and convergent validity testing against objective measures are the next crucial steps.