In order to examine the limitations associated with this technique, a systematic study of the various parameters affecting intensity measurements to determine sound power will be presented. The effect of the finite approximation, phase mismatch, inter‐active intensity, the number of points, and the spatial distribution of the points will be discussed, keeping in mind actual measurement conditions. Somein‐situexperimental studies will be presented. For a typical section of a textile machine, the power measured via ISO 3744 in a semianechoic field compared well with that measured in a semireverberant room using intensity, confirming the validity of the technique. Further, it will be shown that intensity can be used to differentiate the noise emitted by the frame from that of the spindle. A systematic way of including a criterion of precision will be proposed. This leads to a tolerance which gives the degree of confidence on the measurement and which points out clearly unfavorable conditions.