The invention relates to a package workshop quality data processing method. Data of the cigarette physical index comprehensive test board are stored in a socket mode, information such as relevant parameters and states of equipment in the production process is stored in an OPC mode, and data such as equipment maintenance, appearance defects and auxiliary material information are manually input andstored through an MES system. Large deviation is generated on auxiliary results by abnormal data caused by manual operation errors and equipment contingency in the production process, and the data arescreened and cleaned through the rules of the data to obtain quality data meeting the rules. The dispersed data of the storage nodes is integrated through correlation factors. Reduction processing isperformed on the data through redundant storage and irrelevant data. The unstructured stored data is converted into data capable of being directly used through the data, efficient data calling can beachieved through the storage mode of the cloud data center, and an auxiliary decision is made for accurately making the quality of the tobacco package workshop.