Bronchoalveolar lavage has proved a useful research technique for recovering cellular and molecular contents of the lower respiratory tract. Because the recovered fluid is variably diluted, an accurate estimation of molecular and cellular concentrations can only be made if the epithelial lining fluid volume recovered is also known. It has been suggested that smoking may alter epithelial lining fluid volume by reducing clearance or by stimulating production and, thus, affect the interpretation of bronchoalveolar lavage studies. In this study, urea was used as an endogenous marker of epithelial lining fluid volume in a comparison of 26 smokers and 31 nonsmokers. The mean epithelial lining fluid volume recovered from smokers was significantly greater than that of nonsmokers (2.4 +/- 1.40 ml vs 1.2 +/- 0.75 ml, p less than 0.005). The total cellular concentration in the bronchoalveolar lavage fluid in smokers was also greater (94.2 +/- 46 x 10(6) vs 33.9 +/- 21.5 x 10(6) cells per 300 ml lavage), even when corrected for bronchoalveolar lavage volume recovered (63.1 +/- 32.5 x 10(6) vs 24.9 +/- 13.3 x 10(6) cells per 100 ml recovered lavage fluid). This was true for macrophage, lymphocyte and neutrophil cell numbers. However, when corrected for the apparent epithelial lining fluid volume, only the macrophage count remained significantly higher in the smokers compared with nonsmokers (30.66 +/- 20.7 x 10(6) vs 18.21 +/- 8.6 x 10(6) macrophages.ml-1 ELF). In addition, concentrations of albumin and immunoglobulin M (IgM) were significantly lower in smokers after correction for epithelial lining fluid volume. These results highlight smoking as a confounding factor in the interpretation of bronchoalveolar lavage data.(ABSTRACT TRUNCATED AT 250 WORDS)