Clear‐air turbulence (CAT) presents a scientific problem of considerable importance to aviation and to attempts to understand atmospheric processes that must be incorporated in a long‐range numerical prediction model. Present knowledge about CAT is reviewed briefly, and the fact that the requirements of both aviation and atmospheric science for new information necessitate a measurement program whose objective is increased understanding of the physical mechanisms of CAT is emphasized. Specific investigations that should be a part of such a measurement program are presented, and the importance of the energy budget of CAT is stressed both for its scientific value and for its potential as a method of verifying accuracy of measured data. A new approach to determining the required accuracy of turbulence instrumentation, utilizing root‐mean‐square error techniques, is presented. Some aspects of the management of a large measurement program are also con