Based on analyzing the characteristics of fault diagnosis and practical elimination of smart electric, water and gas meter, an abnormity diagnosis and treatment method of smart meters based on association mining is proposed by this paper. First, diagnosis rules are obtained through association mining by this method, the monitored data is then converted to event elements and the diagnostic results are obtained via a rule match, finally, the diagnostic results are confirmed and feedback through the field investigation. The abnormity diagnosis and elimination of closed-loop management is combined by the method, and the accuracy of the abnormity diagnosis is increased by improving the diagnosis knowledge base, and the superiority of the association mining technology is fully exerted.#$NL Keywords: Association mining; Abnormity diagnosis; Closed-loop management; Diagnosis knowledge base