Not all laboratories are alike.

KLUG et al. [1] have reported observer variability within and between two experienced observers of a number of lung function measurements in preschool children. One of the methods assessed was the measurement of airway resistance by the interrupter technique (Rint). In their laboratory, Rint measurements have high interobserver variability, although no systematic bias. In our laboratory interobserver variability of Rint measurements is measured with each new observer to ensure that it is acceptable [2]. Although our within-observer variability is similar to that reported by KLUG et al., the between observer variability is much lower, despite differences in experience. There is no systematic bias. KLUG et al. have shown a variance (2 SDs of the differences between observers measurements) of 0.62 kPa. L-1. s, where ours is 0.14–0.18 kPa. L-1. s (table 1). Our 95% limits of agreement are therefore much narrower. It is misleading to imply that interobserver repeat-ability for Rint is generally poor. Technical aspects of measurement and different criteria for accepting measurements both affect repeatability. Each laboratory should have its own standards for intra-and interobserver repeatability and use these to power proposed studies. of lung function measurements in 2–6 yr-old children.