AIM: To determine if severity assessment after 1 h of treatment is better than assessment at presentation for predicting the requirement for hospital admission for emergency department (ED) patients with acute asthma. METHODS: Prospective, observational study conducted in 36 Australian ED for a 2-week period in 2001 involving patients aged 1-55 years presenting with asthma. Data collected included severity assessment according to the National Asthma Guidelines (Australia) at presentation and 1 h, and disposition. Descriptive analysis was applied. RESULTS: 720 cases were analysed. Patients with 'mild' asthma at either assessment time had a greater than 80% chance of discharge home. Patients assessed as 'severe' at either assessment had a greater than 85% chance of requiring hospital admission, but the 1 h assessment was better at predicting the need for Intensive Care Unit (ICU) admission. For the 'moderate' group, the initial assessment was a poor predictor of the need for admission however those who met the criteria for 'moderate' severity at 1 h had an 84% chance of requiring admission. CONCLUSION: Assessment of asthma severity after 1 h of treatment is better than initial severity assessment for determining the need for hospital admission for patients initially assessed as having 'moderate' asthma and for predicting the need for ICU in patients initially assessed as 'severe'.