Toward　the　non-stationary　random　properties　of　acoustic　leakage　signal,　the　tunable　Q-factor　wavelet　transform　is　applied　to　analysis　of　acoustic　leakage　signal.　Research　is　conducted　to　investigate　the　influence　of　Q-factor　and　decomposition　scale　on　the　extraction　of　leakage　component.　According　to　the　maximum　of　kurtosis,　the　appropriate　parameters　for　the　tunable　Q-factor　wavelet　are　determined.　The　leakage　components　extracted　by　Q-factor　wavelet　transform　are　used　for　leakage　location.　It　is　demonstrated　by　experiments　that　the　leakage　components　can　be　effectively　extracted　by　tunable　Q-factor　wavelet　transform,　and　the　leak　can　be　located　correctly.　The　research　provide　a　potential　tool　for　leak　detection　in　pipe　of　industry.　©　2018　Acta　Acustica.