The impact of ultrasonic power density on changes of heavy metals during sludge sonication was inves-tigated. Results showed that ultrasound could release heavy metals from sludge into the supernatant. There existed an effective power density range of 0.8-1.6 W·ml?1 for the release of the total heavy metal; there was little release below 0.8 W·ml?1 and too high power density was adverse to the release. Furthermore, sonication showed selective release of heavy metal from sludge to the supernatant; copper, cadmium and lead were not released by sonication, while arsenic and nickel were released easily and their release ratio could reach 40%. The effective energy range for each heavy metal was also different that 0.8-1.2 W·ml?1 for arsenic, 0.5-1.6 W·ml?1 for nickel, and 0.8-1.6 W·ml?1 for mercury and chrome. The differences among heavy metal release during sonication might be explained by the different distribution of chemical fractions of each metal in sludge. Such selectivity could be used to control heavy metal release during sludge treatment. The impact of ultrasonic power density on changes of heavy metals during sludge sonication was inves-tigated. Results showed that ultrasound could release heavy metals from sludge into the supernatant. There existed an effective power density range of 0.8-1.6 W · ml -1 for the release of the total heavy metal; there was little release below 0.8 W · ml -1 and too high power density was adverse to the release. were not released by sonication, while arsenic and nickel were released easily and their release ratio could reach 40%. The effective energy range for each heavy metal was also different that 0.8-1.2 W · ml -1 for arsenic, 0.5-1.6 W · ml? 1 for nickel, and 0.8-1.6 W · ml? 1 for mercury and chrome. The differences among heavy metal release during sonication might be explained by the different distribution of chemical fractions of each metal in sludge. Suchfunction be used to control heavy metal release during sludge treatment.