In　previous　work　we　have　shown　that　recycling　pre-sonicated　drinking　water　treatment　sludge　(DWTS)　could　improve　coagulated　water　quality.　Here,　the　removal　of　naturally　occurring　organic　matter　of　source　water　was　further　optimized　using　response　surface　methodology　(RSM)　with　Box-Behnken　Design　(BBD).　The　four　variables,　i.e.,　volumetric　recycling　ratio　of　DWTS,　energy　density,　ultra-sonication　time　and　duty　cycle　in　an　experimental　jar　test　of　ultrasound　assisted　flocculation-coagulation　were　optimized.　All　the　variables　showed　a　significant　effect　on　dissolved　organic　carbon　(DOC)　removal　of　source　water　(p　＜.05),　of　which　the　duty　cycle　had　a　stronger　effect　on　the　removal　performance　compared　to　the　other　independent　variables.　The　predicted　optimal　DOC　removal　rate　was　36.94%,　and　this　matched　well　the　observed　performance　of　36.54　+/-　0.56%,　obtained　by　ultra-sonicating　the　sludge　prior　to　recycling　using　a　power　input　of　1.015　W/mL,　an　ultra-sonication　time　of　9　min　50　s,　and　a　duty　cycle　of　80%,　while　the　volumetric　recycling　ratio　of　DWTS　was　5.8%.　The　natural　organic　matter　fractions　in　the　coagulated　water　samples　indicated　that　recycling　sonicated　DWTS　that　had　been　washed　prior　to　recycling　in　order　to　remove　solubilized　extracellular　polymers　could　enhance　removal　of　hydrophobic　acids　and　3-30　kDa　fractions,　but　this　treatment　increased　the　presence　of　substances　with　molecular　weight　＜　3　kDa.　Humic-like　substances　were　effectively　removed　while　tyrosine-like　substances　could　be　enriched.　Sludge　samples　(raw　DWTS,　sonicated　DWTS,　sludge　formed　by　recycling　raw　DWTS,　and　sludge　formed　by　recycling　sonicated　DWTS　without　solubilized　extracellular　organics)　were　characterized　by　XRF,　X-ray　diffraction　patterns　and　FE-SEM-EDS　to　reveal　possible　physical　characteristics　that　could　be　related　to　the　DOC　removal　performance.