The invention discloses a cigarette tobacco cutting process tobacco flake preprocessing stage on-line monitoring and fault diagnosis method. As for key equipment in a tobacco flake preprocessing stage such as TB-L loosing and conditioning equipment and KAS primary feeding equipment, the problem of incapability of realizing accurate monitoring and fault diagnosis which is caused by a large number of batches, slow time varying, uncertain operation time and diversity of products may be brought about, while, with the method of the invention adopted, the problem can be solved. The method includes the following steps that: operating data of batch, time and attribute three-dimensional characteristics are unfolded according to attribute directions through process characteristic analysis, so that the problem of unequalness of the length of data of different batches can be solved; a monitoring model of each kind of flake product trademark is established respectively through adopting a principal component analysis (PCA) and based on a multi-model structure, and the T<2> and SPE statistics and control limits thereof of monitoring models of different trademarks are calculated in an off-line manner; process operating data of a tobacco flake preprocessing stage are acquired in an online manner, and corresponding monitoring models are called according to product trademarks, so that the T<2> and SPE statistics can be calculated in an online manner; and when any index exceeds the control limit of a normal operating area, a contribution chart method is adopted to perform fault diagnosis.