Research conducted in 10 cities assessed long-term pediatric asthma outcomes from a peer teaching intervention for physicians to improve their asthma-related clinical and counseling skills. Hypotheses were better outcomes for patients, symptom reduction, less health care use, and enhanced view of the physician. Peers trained 53 intervention group pediatricians (seeing 418 patients); 48 pediatricians (seeing 452 patients) were controls. Patients provided baseline and 2-year follow-up data, collected by telephone interview and from medical records. Intent-to-treat analyses used Poisson regression and general estimation equations. Treatment physicians' patients gave them higher performance ratings ( P = .02). Patients had fewer sleep disruptions from asthma symptoms ( P = .03). Those with baseline health care use had fewer ED visits ( P = .005), hospitalizations (P = .03), and urgent office visits (P = .001), and they made fewer phone calls to the doctor's office (P = .02). Treatment physicians spent no more patient visit time than control physicians. Peer training increased patient's positive views of clinician's performance and reduced children's symptoms and health care use up to two years post program.