The usefulness of single breath nitrogen test (SBN2) was evaluated in a cross-sectional epidemiologic survey on a general population sample (n = 3,289) of North Italy. Each subject was submitted to CNR standardized questionnaire and to lung function testing using automated equipment (Hewlett-Packard 47804S). Acceptable closing volume (CV) and slope of alveolar plateau (DN2%/L) tracings were performed by only 1,370 and 1,982 subjects respectively, in comparison with the 2,638 diffusing capacity and the 3,079 forced vital capacity acceptable maneuvers. Prediction equations were computed on normal subjects for CV indices, with the exception of DN2%/L: they were similar to those found in other studies. Significant differences among smoking categories were found for all the SBN2 parameters in males, but only for DN2%/L in females. The latter was also able to distinguish either subjects with airway obstruction from those without it or symptomatic from asymptomatic individuals, but DN2%/L did not give more information than Vmax75, a sensitive index of flow volume curve. Our results suggest that the place of SBN2 in large scale epidemiologic testing has not been justified.