We read with interest the article by Berbescu et al (May 2006),1 who attributed a role for transbronchial lung biopsy (TBLB) in the diagnosis of usual interstitial pneumonia (UIP). Their conclusions raise serious issues, aside from the potential bias in this unblinded retrospective study. Irrespective of operator expertise, TBLB has inherent sampling errors, particularly in patients with established lung fibrosis. Small specimen size makes TBLB a “histopathologist’s nightmare,” with difficulty in distinguishing different patterns within the spectrum of diffuse parenchymal lung diseases; patients may have overlapping histologic features. Berbescu et al1failed to mention sample size. Adequate biopsy size, ideally a 4-cm maximum diameter when inflated, and a depth of at least 1 to 1.5 cm,2are critical to identify potential prognostic markers, such as the degree of alveolar space granulation tissue deposition and the extent of early connective tissue formation within the fibroblastic foci, in patients with UIP; such factors may also impact on treatment outcome.3–4 An additional inevitable crush effect, a failure to penetrate beyond the peribronchial sheath, and friable tissue disintegration preclude proper histologic assessment.