Although many physicians and patients want to replace low dose inhaled steroids with montelukast in controlled patients with mild asthma, there are insufficient data to predict montelukast failure. A study by Drummond et al (J Asthma 2011 Oct 27 [Epub ahead of print]) provides a piece of the puzzle. Using the 165 participants in the Leukotriene or Corticosteroid or Corticosteroid-Salmeterol Study (LOCCS) trial who were stepped down from low-dose ICS to montelukast, the authors attempted to predict the risk of montelukast treatment failure during step-down. Characteristics independently associated with montelukast treatment failure included early asthma onset (<10 years), need for steroid burst in the last year, and low pre-bronchodilator FEV1. They constructed a montelukast failure index that may prove to be helpful for clinical practice, but it needs further validation.