Abstract: An adaptive recursive partial least squares(RPLS ) monitoring scheme is proposed to improve the time-variant tracking power of the statistical quality monitoring (SQM) systemsWhen new samples are obtained, this new method updates the monitoring model and the Q statistic control limit (Qα) on the basis of the improved RPLS algorithmsThus it could overcome the shortage of the traditional fixed SQM successfullyThe theoretical findings were fully supported by the application performed on the rubber mixing process in a large-scale tire plant in east ChinaIt was shown that the compounds quality is improved remarkably and the false alarm frequency was reduced significantly.