OBJECTIVES Poor adherence to ICS among individuals with asthma is common and associated with increased asthma symptoms and acute health care utilization; diminished quality of life; and increased health care costs. We sought to determine if a brief, automated telephone intervention using speech recognition (SR) technology improved ICS adherence among adults with asthma. METHODS PEANUT is an ongoing clinical trial involving 14,064 members of a health maintenance organization, aged >18 years, taking medication for the management of asthma. Participants were randomized to usual care (UC) or the 18-month SR arm. SR participants taking ICS receive monthly refill reminder calls, as needed, based on dispensing data derived from the electronic medical record. The calls were designed using qualitative methods and provide educational information, explore barriers to adherence, and offer transfer to an automated refill line as appropriate. The primary outcome is the continuous measure of medication adherence (CMA), a measure of medication days dispensed relative to total observation days. Preliminary results through the first 8 months are presented. RESULTS Of over 10,000 calls attempted so far, we successfully reached the target participant 45% of the time and left a voice message on an additional 40% of calls. CMA scores were significantly higher for SR than UC participants (0.48 vs.0.45, p<0.01), with the strongest effects (a five percentage point net differential) in those aged >60 yrs. CONCLUSIONS Early results suggest a small, significant intervention effect. Because small changes in adherence on a population basis can have important public health benefits, this study supports the potential value of SR-based adherence interventions for asthma and other chronic diseases.