BACKGROUND Emphysema is currently defined as "a condition of the lung characterised by abnormal, permanent enlargement of the airspaces distal to the terminal bronchiole, accompanied by destruction of their walls, and without obvious fibrosis." The functional and morphological changes that occur in emphysema have largely been attributed to changes in alveolar elastin rather than in collagen. A study was performed to determine whether the amount of collagen in the alveolar wall changes with age in the lungs of non-smokers and of smokers with different types of macroscopically defined emphysema in relation to a microscopic measurement of lung structure. METHODS Total alveolar wall collagen was measured (as hydroxyproline) in known volumes of distended lung tissue (by reverse phase high pressure liquid chromatography) in the lungs of non-smokers (n = 23) and in regions sampled away from emphysematous lesions in the lungs of 36 smokers (four with no emphysema, 13 with centriacinar emphysema (CAE), nine with panacinar emphysema (PAE), and 10 with a mixture (MIX) of both PAE and CAE). Mean lung airspace wall surface area per unit volume (AWUV) was calculated from at least six random blocks per lung and on histological sections immediately adjacent to those prepared for collagen measurement with a rapid scanning device (fast interval processor). RESULTS In non-smokers there was no significant correlation between the amount of collagen in the alveolar wall tissue and either mean lung AWUV or increasing patient age when amounts of collagen were expressed either per unit volume of distended lung (40 mm3 sample) or per unit surface area of airspace wall tissue. Smokers without emphysema had similar amounts of collagen to non-smokers. Lungs with PAE and MIX, but not CAE alone, contained significantly more collagen than normal when expressed per unit volume of airspace wall tissue whereas all groups, including CAE, contained significantly raised amounts of collagen when expressed per unit surface area. CONCLUSIONS There is no significant age related change in the collagen content of the lungs of non-smokers which suggests that, as AWUV is lost with age, the main collagenous framework is maintained. However, in smokers with emphysema there is a loss of airspace wall tissue in regions remote from the macroscopic lesions that is accompanied by a net increase in collagen mass. The greater accumulation of collagen in MIX lungs than in CAE lungs suggests a greater degree of structural damage, indicative of an alternative pathogenetic mechanism operating between the different types of emphysema. Our results suggest an active alveolar wall fibrosis in emphysema as a consequence of cigarette smoking. It is suggested that the definition of emphysema may require further revision to include such change.