T measurement of air speed of airplanes by means of instruments attached to them is often very inaccurate. The sources of error are numerous and not all of them are easily avoided. It is the purpose of this paper to discuss the results of experience gained in numerous flight tests conducted at the laboratory of the N.A.C.A. at Langley Field. Two aspects of the problem are considered, one being the measurement of air speed with a high degree of precision using special test equipment, and the other being the measurement of air speed with fair precision for normal flying. The latter aspect of the problem includes consideration of the provision of a satisfactory reference pressure for the altimeter. The use of a pitot-static head is assumed throughout.