OBJECTIVES: Our aim was to evaluate the construct validity of generic and disease-specific measures of health in chronic obstructive pulmonary disease (COPD) with a specific focus on the relative ability of the measures to discriminate between clinically defined stages of COPD.  METHODS: We analyzed data collected from 120 patients with a diagnosis of COPD in a Veterans Affairs hospital.  Patients self-completed 2 generic measures (EQ-5D and SF-36), and the disease-specific measure St. George’s Respiratory Questionnaire (SGRQ).  Known groups comparisons for the SGRQ, EQ-index, VAS, MCS-12, and PCS-12 scores were conducted by grouping patients according to Global Obstructive Lung Disease (GOLD) stages based on forced expiratory volume in one second (FEV1).  Pearson’s correlation was used to assess the associations between clinical and HRQL measurements.  Relative efficiency ratios (REs) were used to assess the statistical efficiency of each summary score in discriminating among levels of clinical severity of COPD.   RESULTS: Mean PCS-12, EQ-VAS and SGRQ total scores significantly declined with stage of severity (p <0.05).  The SGRQ was more discriminative than EQ-5D and SF-36 in the severe stages.  The generic summary scores demonstrated trivial correlations with FEV1 (r<0.2) whereas SGRQ showed a moderate correlation (r=0.43).  The SGRQ exhibited a stronger correlation with Borg dyspnea score (r>0.5) than the generic summary scores.  Using SGRQ as the reference, PCS-36, MCS-36, EQ index, EQ-VAS all had REs of <1. CONCLUSIONS: The SGRQ demonstrated greater ability to discriminate among different levels of severity stages of COPD and is more strongly correlated with clinical measures of COPD than generic measures of health, an important consideration when selecting measures in studying patients with COPD.