We investigated the utility of transbronchial lung biopsy in allograft evaluation, particularly with reference to rejection-mediated bronchiolitis obliterans in 105 transplant recipients surviving 3 months or more (29 single lung, 76 double lung). A sensitivity and specificity of one transbronchial biopsy procedure with an average procurement of 7.6 tissue fragments was 17.1% and 94.5%, respectively, according to results obtained from biopsies carried out after 2 months transplantation in 29 patients with confirmed rejection-mediated bronchiolitis obliterans and 61 patients known not to have the disease. The predictive value of a positive procedure for the presence of disease was 65.5% and that of a negative procedure for the absence of disease was 65.2%. Similarly, the sensitivity and specificity for the finding of bronchiolar mural fibrosis were 18.5% and 85.3%, respectively; and, for lymphocytic bronchiolitis, the values were 1.9% and 100%, respectively. We conclude that, although the sensitivity of transbronchial biopsy and the predictive value of a procedure which shows microscopic bronchiolitis obliterans are low, attempts should be made to support a clinical diagnosis of bronchiolitis obliterans through biopsy, given the problems inherent in achieving an early and certain diagnosis according to clinical criteria alone.