BACKGROUND: Evidence-based guidelines exist to guide inpatient management of chronic obstructive pulmonary disease (COPD) exacerbations, but we do not know how well these recommendations are adhered to. AIMS: The aims of this study were: (i) to examine concordance with evidence based guidelines for inpatient management of COPD and (ii)to address deficiencies in compliance with guidelines by feedback of audit results and distribution of an education package. METHODS: Retrospective chart reviews were performed using a data collection tool based on current guidelines. Forty-nine consecutive COPD admissions were audited, and results presented to medical staff. An education package was distributed directly after the presentation. One month later,35 consecutive separations were reviewed. Concordance with recommendations supported by the highest level of evidence was calculated. RESULTS: Data were complete for 84 cases. Concordance rates ranged from 0 to 100%. Apart from initiation of systemic steroids (80-83%)and avoidance of intravenous aminophylline (100%),concordance rates were less than 60%. The only significant improvement post-intervention was for steroid duration (10 vs 29%,95% confidence interval for difference (-36.2, -1.8)). CONCLUSION: Recommendations for steroid initiation and avoidance of aminophylline are well adhered to. Concordance rates for other recommendations were generally less than 60%. Concordance with recommendations for steroid duration was significantly improved by our intervention. The findings suggest that to facilitate evidence-based practice, alternative interventions should be evaluated.