Medical and surgical advances have made lung transplantation a feasible therapy for end-stage lung disease. Fiberoptic bronchoscopy with bronchoalveolar lavage (BAL) and transbronchial lung biopsy (TBBx) is an accepted technique for detecting clinically evident rejection and infection in the allograft of symptomatic recipients. The role of TBBx and BAL in managing asymptomatic recipients is less defined. We retrospectively examined the role of bronchoscopy with TBBx and BAL in 1124 bronchoscopy procedures that were performed on 161 lung transplant recipients between January 1, 1988, and December 31, 1993. Bronchoscopy was performed when there was a change in the recipient's clinical condition, to assess the response of the allograft to a prior therapy, and under a surveillance protocol for detecting asymptomatic rejection or infection. Surveillance bronchoscopy was performed according to the following schedule: 10-14 days after transplantation, every 3 mo during the first year, every 4 mo during the second year, and at 6-mo intervals thereafter. Surveillance bronchoscopies were defined as procedures where the physician felt that there was no infection or rejection in the allograft on the basis of a standardized clinical evaluation, which excluded the results of the TBBx and BAL. We compared the clinical impression recorded by the physician on the day of the procedure with the final diagnosis determined after the results of the TBBx and BAL were known. We found unsuspected rejection and/or infection that required therapy in 25% (90/355) of all surveillance bronchoscopy procedures. Most episodes (61/90, 68%) of unsuspected rejection and/or infection occurred in the first 6 mo after transplantation.(ABSTRACT TRUNCATED AT 250 WORDS)