Objective. This article describes an American Institute of Ultrasound in Medicine-sponsored intercomparison of the results of acoustic power measurements performed by several laboratories. Methods. Two primary calibration techniques, namely, planar scanning and radiation force balance, were used in the frequency range typical of that in which sonographic imaging devices operate. The same reference source, the National Institute of Standards and Technology (Gaithersburg, MD) standard ultrasonic power source, capable of producing acoustic fields in the frequency range from approximately 1 to 21 MHz, was circulated to 3 laboratories. Results. The results of the calibrations indicate that the overall uncertainty in acoustic power measurements depends on the target and the measurement method. In the case of radiation force balance measurements with an absorbing target, the largest discrepancy between the available National Institute of Standards and Technology-calibrated results and the reported data was 10.6% at approximately 2.5 MHz. At higher frequencies, beyond 10 MHz, the largest discrepancy reported with an absorbing target was 8.4%. For a reflecting target, the largest discrepancies were 16.2% at approximately 3.7 MHz and 15.4% at about 10 MHz. The largest discrepancy identified for the planar scanning technique below 10 MHz was 7.4% at 3.7 MHz. Conclusions. The results obtained suggest that an absorbing target may be preferable for acoustic power measurements with radiation force balance. In a group that consists of 2 research laboratories and 1 manufacturer, the power measurements agreed within 16%. Key words: planar scanning technique; radiation force balance; ultrasonic hydrophones; ultrasonic transducers.