INTRODUCTION: There is no universally accepted measurement scale to assess breathlessness in adult palliative care patients. This significantly hampers clinical practice and research into effective interventions. The aim is to systematically identify and appraise breathlessness measurement scales, which are validated for use in palliative care or which show potential for use. METHODS: We undertook systematic searches of electronic databases (Cochrane databases 2005, MEDLINE 1966-2005, OLDMEDLINE 1950-1965, EMBASE 1980-2005, PsycINFO 1872-2005, AMED 1985-2005, CINAHL 1982-2005, SIGLE 1980-2005) with follow-up searches (reference lists of included papers, hand-searches of relevant journals). The basic search strategy was 'breathlessness (etc.) AND measurement (scales, validation etc.) AND palliative care/cardiac failure/respiratory disease/ neoplasm etc.', modified for each database, without language restriction. Patient-based scales with evaluations of at least two psychometric characteristics were included. Exercise-based tests were excluded. Scales were appraised with particular emphasis on construct validity and responsiveness. RESULTS: We identified 29 scales: six to measure breathlessness severity, four to assess breathlessness descriptions, and 19 to measure functional impact of breathlessness. SEVERITY: The Numeric Rating Scale (NRS) and modified Borg Scale have been evaluated in COPD (the NRS has also been evaluated in cancer). Both require further assessment of responsiveness and test-retest reliability over time intervals relevant to palliative care. Visual Analogue Scales have also been evaluated, but require larger sample sizes than NRS for evidence of intervention effectiveness. DESCRIPTIONS: The Japanese Cancer Dyspnoea Scale (CDS) has been evaluated in patients with cancer, but requires further assessment of construct validity and responsiveness. FUNCTIONAL IMPACT: The Chronic Respiratory Questionnaire dyspnoea subscale (CRQ-D) has been evaluated in chronic lung diseases and heart failure; the MND Respiratory Scale is similar. CRQ-D has face and construct validity, test-retest reliability and responsiveness, and shows promise for palliative care. CONCLUSION: The NRS, modified Borg, CRQ-D and CDS appear most suitable for use in palliative care, but further evaluation is required before adopting any scale as standard. This review has been registered with the Cochrane collaboration and will be published and updated as a Cochrane review.