Re-examination of the pathologic and clinical features of the entities traditionally classified under the heading "pulmonary angiitis and granulomatosis" indicates that there is little advantage in retaining this artificial category and that these entities should be considered variants of diseases to which they are actually related. Wegener's granulomatosis and allergic angiitis and granulomatosis appear to be examples of true systemic vasculitides in which the lung is a predominant but not the only or even the most important site of involvement. Wegener's granulomatosis may manifest with involvement limited to lung, a form that has been called limited Wegener's; however, many or most such cases progress to classic disease involving kidney and often upper respiratory tract. Similarly, Wegener's granulomatosis may present with disease limited initially to the upper respiratory tract (a form of midline granuloma); this process may also spread to involve lung and kidney. It seems unlikely that limited Wegener's is truly a separate disease category. Evaluation of the pathologic and clinical features of necrotizing sarcoid granulomatosis indicate that it very much resembles ordinary sarcoid in most histologic features, in the nature of extrapulmonary involvement, and in its clinical course and that it probably corresponds to the clinical--radiographic entity of nodular sarcoid. Lymphomatoid granulomatosis appears to have little relationship to the other members of the angiitis and granulomatosis group; its behavior and histologic features are those of a lymphoproliferative disorder that in most cases is or becomes histiocytic lymphoma. Some cases of so-called benign lymphocytic angiitis also fall into this category; the remainder appear to represent a variety of completely unrelated pathologic processes. Last, bronchocentric granulomatosis is most commonly one of the histologic manifestations of allergic bronchopulmonary aspergillosis, although it is likely that other agents or processes produce the same histologic pattern. Although the presence of a common set of pathologic features makes the concept of angiitis and granulomatosis attractive from a morphologic point of view, there is minimal clinical similarity among them, and these diseases appear to be totally separate entities.