The uncertainty of measurement results of air velocities in air-conditioned rooms generally is very high. A considerable part of the measurement errors are caused by the evaluation methods.

It is shown that the accuracy of results can be ameliorated by using suitable statistical methods without increasing the effort. In this case binomial distribution gives better results than Gaussian distribution.

To optimize the duration of measurements the most favorable reading interval must be determined. This time interval can be defined by the autocorrelation function of air velocity.