Objectives: To determine whether well trained lay people could deliver asthma self-management education with comparable outcomes to that achieved by primary care based practice nurses. Design: Randomised equivalence trial. Setting: 39 general practices in West London and North West England. Participants: 567 patients with asthma who were on regular maintenance therapy. 15 lay educators were recruited and trained to deliver asthma self-management education. Intervention: An initial consultation of up to 45 min offered either by a lay educator or a practice based primary care nurse, followed by a second shorter face to face consultation and telephone follow-up for 1 year. Main outcome measures: Unscheduled need for healthcare. Secondary outcome measures: Patient satisfaction and need for courses of oral steroids. Results: 567 patients were randomised to care by a nurse (n = 287) or a lay educator (n = 280) and 146 and 171, respectively, attended the first face to face educational session. During the first two consultations, management changes were made in 35/146 patients seen by a practice nurse (24.0%) and in 56/171 patients (32.7%) seen by a lay educator. For 418/567 patients (73.7%), we have 1 year data on use of unscheduled healthcare. Under an intention to treat approach, 61/205 patients (29.8%) in the nurse led group required unscheduled care compared with 65/213 (30.5%) in the lay led group (90% CI for difference -8.1% to 6.6%; 95% CI for difference -9.5% to 8.0%). The 90% CI contained the predetermined equivalence region (-5% to +5%) giving an inconclusive result regarding the equivalence of the two approaches. Despite the fact that all patients had been prescribed regular maintenance therapy, 122/418 patients (29.2%) required courses of steroid tablets during the course of 1 year. Patient satisfaction following the initial face to face consultation was similar in both groups. Conclusions: It is possible to recruit and train lay educators to deliver a discrete area of respiratory care, with comparable outcomes to those seen by nurses. Trial registration number: NCT00129987.