Growing numbers of plant engineers, energy managers, motor manufacturers, and even instrument suppliers are advocating the performance of simple and reliable methods of testing motor efficiency in the field, at a user's plant, as a basis for economic studies of the benefits of higher-efficiency motor designs. It is argued that although these tests are often simple, their accuracy leaves much to be desired. Experimental evaluation of motor losses is quite complex. Results can vary widely even under the best of factory conditions. A field test, therefore, should not be considered a proof test. The stray loss problem, effects of voltage variation, and instrumentation are discussed. >