A test procedure has been developed for sound insulation measurements using the sound intensity method. The basis of the procedure is from international standards ISO 140-3 and ISO 9614-1. For specimens that are sound absorbing on the receiving room side, the transmitted intensity will be underestimated if the extraneous sound intensity "caused by flanking or sound leaks… is high. A flanking ratio check is included in the test procedure to maintain accuracy within desired limits. The consideration of the field nonuniformity indicator, F 4 , in the test procedure was also found necessary. Seals at the edges of movable components of sound isolating structures, for example doors, are often poorly designed and manufactured. Sound leaks increase the value of the field indicator F 4 because of inhomogeneous sound transmission. Improving the seal of finished products is not possible in the test situation. For F 4 to fall in the acceptable range the number of measurement points may have to be increased enormously. The test procedure can be applied both to laboratory and field measurements. The main advantage of the new sound intensity method is the ability to locate sound leaks in the test object. This allows a customer-friendly mechanism for product development. Another advantage is that the laboratory facility can be much lighter than for the conventional method while maintaining the same limit for the maximum measurable sound reduction index. An automatic two-dimensional robot scanning system was used for exact positioning of the intensity probe. © 1997 Institute of Noise Control Engineering. †S0736-2501"97…00502-X ‡