To examine the advantage of hot-wire anemometer for clinical use, we have checked two types of this tools with respect to reliability and validity. It was found that electronic suppression of noise caused a distortion of the measurements. Furthermore changes of transducers were also responsible for deviations from true values. We require of the manufacturer to indicate the threshold of perception and the coefficient of variation for repeated measurements with several transducers. We recommend a simple rule which permits an estimation of the limits of reliable measurements for clinical use depending on the threshold of the equipment and on the parameters of ventilation.