The aim of this work was to describe the results of a simple quality control procedure for the flow-volume curve adopted in a multicentre epidemiological study (PEACE). In 14 centres, 8-15 individuals (n = 157) performed forced vital capacity (FVC) manoeuvres following a standard protocol with both the local spirometer/pneumotachograph and a portable spirometer (i.e. the 'reference instrument' for this study). Deviances of measurements were assessed by computing the differences (delta) between the former and the latter, the ratios of such differences on portable spirometer values (delta %) and the coefficients of variation (CV). The portable spirometer yielded lower mean AFVC and deltaFEV1 (forced in 1 sec) than local instruments (except for two and four centres, respectively). In most instances, differences were statistically significant. Absolute mean A%FVC ranged from 4.9-18.2%, while delta%FEV1 ranged from 2.3-18.5%. The Bland and Altman analysis showed a good agreement between the portable and local instruments, except for two centres, where a systematic trend towards higher individual absolute deltaFVC and deltaFEV1 was observed. The overall variability, assessed by CV, was within 6.2% and 5.1% for FVC and FEV1, respectively: it was similar to other quality control studies ranging from 2.0-5.5% for FVC and 2.2-5.8% for FEV1. Our results point out the importance of performing interlaboratory comparisons as a quality control procedure in multicentre epidemiological studies on lung function, and of stimulating manufacturers to extend the accuracy and precision of the instruments.