OBJECTIVES: Inhaled corticosteroids (ICSs) and long-acting beta-agonists (LABAs) have been recommended for patients with moderate to severe persistent asthma.  This study quantified medication possession ratios (MPRs) for those asthmatic patients and assessed the association between severe asthma exacerbations (SAEs) and degree of medication compliance.     METHODS: The primary data source was the MarketScan Multistate Medicaid database from 2002 to 2007.  A total of 299,917 patients (ages 0-64) with moderate or severe asthma were selected.   Drug use was tracked for ICSs, LABAs, ICS/LABA combination drugs, short-acting beta agonists leukotriene modifiers, and oral and IV steroids. An ICS-and-LABA MPR was calculated as the sum of days’ supply for ICS and LABA drugs divided by the number of follow-up days during the first year after the patient’s asthma index date. Logistic regressions were used to assess the risk of SAEs associated with degree of medication compliance. RESULTS: There were 21,852 (7.3%) patients with severe persistent and 278,065 (92.7%) patients with moderate persistent asthma.  Within 12 months after asthma index date, average MPRs were 0.23 (median 0.14) for ICSs and LABAs and 0.66 (median 0.46) across all asthma medications.  After controlling for covariates like age, gender, race, and comorbidities, the odds ratios of an asthma-related emergency department (ED) visit was 0.81 (0.79-0.84) for a higher ICS-and-LABA MPR, but the odds ratio of asthma-related hospitalization or intubation was 1.56 (1.50-1.62).   Other key risk factors (p<0.0001) included being African American, having an alcohol/substance abuse disorder, pregnancy, obesity, and upper respiratory tract infection. CONCLUSIONS: A higher ICS-and-LABA MPR appears to be associated with a lower risk of asthma-related ED visits, but higher risk of asthma-related hospitalization or intubation.  Certain patients with asthma, such as pregnant adolescents and African Americans, seem particularly vulnerable to SAEs.