The invention relates to a health evaluation method based on full-life-cycle data of an electric energy meter. According to full-life-cycle data of the electric energy meter, the electric energy meterserves as a node, data requirements are integrated, data cleaning and feature screening are completed, fault factors and feature attributes of the electric energy meter are extracted through a data mining algorithm, health scoring is conducted on the electric energy meter through a comprehensive evaluation algorithm, and health grades are divided. The method is deeply ploughed in the power industry based on a big data concept, power grid internal data and external environment data are fused, influence analysis is performed on health-related factors of the electric energy meter, an electric energy meter fault prediction model is constructed, the operation state of the electric energy meter is evaluated according to the mounting and dismounting times, the event reporting times, the acquisition success rate, the acquisition electric quantity abnormality rate, the operation duration, the first inspection error and other dimension data of the electric energy meter, a health evaluation system of the electric energy meter is established, and the health degree of the operation state of the electric energy meter is given.