Some diseases are inherently more difficult to diagnose with certainty than others. Most diseases are easier to diagnose when the clinical presentation fits the paradigm that we learn as a function of our training and experience. The opposite is also true—if the presentation is atypical, the diagnosis becomes much more difficult. One such example is a pulmonary embolus that complicates severe chronic obstructive lung disease, particularly if the patient is being supported by mechanical ventilation.1 Pulmonary physicians are likely to be a little surprised at the data presented by Judson and his colleagues in the current issue of CHEST (see page 406), in which they detail the difficulty encountered by physicians in making a definitive diagnosis of sarcoidosis. This was especially true when patients with sarcoidosis had pulmonary symptoms as the presenting problem; however, when skin manifestations were the first thing noticed by the patient and the physician, the tissue diagnosis of sarcoidosis was made significantly quicker. Does this mean that it is easier to recognize and adequately categorize the skin lesions as typical of sarcoidosis? Not necessarily, but the authors point out that skin biopsies are likely to be done if the skin lesions do not respond to the initial treatment. That makes sense to me; I struggle with most dermatologic diagnoses, and I am quick to call a consultant to help me for all but the simplest of dermatologic problems. And dermatologists proceed to sample such lesions, often at the time of the first encounter.