OBJECTIVES: Despite the known importance of regular use of inhaled corticosteroids (ICS) for asthma control, there is a dearth of adherence literature from India. The study aims were to (i) evaluate patients’ self-reported adherence to ICS therapy, (ii) identify possible reasons for non-adherence, and (iii) identify possible correlations between ICS adherence, medicine beliefs and socio-demographic factors of asthma patients.  METHODS: Adults with previously diagnosed asthma (n=200) presenting to the emergency room (March 2009-December 2009) of a public chest hospital in Delhi for asthma exacerbation completed an interviewer-administered questionnaire on socio-demographics, clinical history, self-reported adherence, beliefs about causes of disease, medicine beliefs and medication adherence after stabilization of their condition.   RESULTS: Study sample had 51.5% patients in the age group of 30-40 years, 54.0% females, 10.0% illiterate, and 62.5% patients earning less than INR 10,000/month ($185/month). The mean duration of registration with the study hospital was 5.4±4.4 years and all were prescribed ICS treatment. Self-report on adherence: 49% took medicines even without symptoms; 91.0% reported they forgot to take their medicine “some or lot of times”; 84.0% avoided medicines “some of the times”. Important reasons for avoiding medicines were no symptoms (59.5%), cost (34.5%), fear of getting dependent (29%), side effects (17%) and social inhibition (14.5%). Correlation between self-reported adherence and demographic factors, such as age, sex, education and income was calculated. The commonly endorsed causes for asthma were pollution (33.5%), poor medical care in the past (15.5%), heredity (8.5%), cold climate (7.5%), and diet (5.5%). Non-adherent behaviors were associated with doubts about the necessity of medication and concerns about its potential side effects with long-term use (r=-0.299, p<0.001).  CONCLUSIONS: Adherence to ICS therapy is poor and many factors modulate adherence to therapy. These findings lend preliminary support for an extended self-regulatory model of treatment adherence, which incorporates beliefs about treatment.