Critical　peak　pricing　(CPP)　is　an　important　means　of　demand　response　in　electricity　markets.　As　a　flexible　tariff　mechanism,　making　rational　CPP　implementation　strategies　and　selecting　suitable　critical　days　are　crucial　to　its　successful　implementation.　Based　on　an　analysis　of　tariff　scheme　and　implementation　process　of　CPP,　customer　response　to　CPP　is　described　by　a　price　elasticity　matrix　of　demand,　and　electricity　prices　are　forecasted　by　a　hybrid　electricity　price　model.　Furthermore,　a　CPP　decision　model　which　considers　the　interests　of　both　customers　and　load　serving　entity　(LSE)　is　introduced　and　then　solved　by　0-1　integer　programming.　Numerical　results　are　finally　used　to　prove　the　effectiveness　of　the　proposed　model,　which　is　beneficial　to　saving　customer　electricity　bills,　reducing　electricity　purchase　cost　of　LSE,　hedging　against　electricity　purchasing　risk　for　LSE　in　wholesale　market,　and　realizing　multi-party　win.