During a standard flexible bronchoscopy procedure, the bronchoscopist has several possibilities for obtaining material from a suspicious lesion, such as biopsy, brushing, washing, and TBNA. For endoscopically visible lesions, the easiest and most successful technique is taking biopsy specimens of the suspicious area under visual control with a minimum of five biopsy specimens recommended. Especially in submucosal tumors that are covered by normal mucosa without signs of tumor, it might take a few biopsies before it becomes possible to obtain material from the submucosal lesion. Whether brushes and/or washings have additional value in this situation is unclear. Another problem is the necrotic debris often found on top of a malignant lesion. Several biopsies might be needed to obtain viable tumor tissue. In a number of patients with this problem, rigid bronchoscopy enables the bronchoscopist to take larger biopsy samples, thus overcoming the covering layers of still normal mucosa or necrotic debris. For nonvisible lesions, taking a biopsy or brush specimen under visual control is almost impossible. Adding fluoroscopy improves this, especially in very experienced hands, but still biopsy and brush samples from several peripheral lesions may be negative. In these situations, one might expect additional value from washing of the part of the lung with the suspicious lesion in it. In the study by van der Drift and coworkers in this issue of CHEST (see page 394), the authors performed a prospective study in a large number of patients with what, in retrospect, were all proven, malignant tumors in the lung. In all patients, biopsy, brushing, and washings were performed, and the effectiveness of washing before and after biopsy and brushing was determined. Furthermore, the authors performed a detailed cost analysis of the different procedures, and calculated the costs of all kinds of combinations and scenarios with washings and brushings as the variable. Within this cost analysis, they took into account the costs of additional diagnostic procedures if bronchoscopy, using all possible ways of tissue procurement, failed to come to a final diagnosis. The following conclusions of this study are of value for the daily practice: when a washing is done is not important, just do it; and perform all techniques for obtaining material for the pathologist but work cost-effectively and ask him to investigate brushings and washings if the biopsy specimen is not diagnostic. Despite these extensive sampling procedures, the diagnostic yield in invisible tumors is still slightly > 50%. This indicates that other approaches have to be investigated to prevent unnecessary expensive additional investigations.