As for the approximate single-peak sequences or fluctuating sequences under saturated condition in human body pharmacokinetics,a novel self-memory GM(1,1)power coupling prediction model is put forward to expand the applicable range of grey prediction model and promote its predictive performance.It has achieved organic coupling of the self-memory principle of dynamic system and conventionalGM(1,1)power model.The conventionalgrey prediction model's weakness as being sensitive to initial value can be overcome by the self-memory principle.As shown in the illustrative example of serum concentrationprediction,the proposed coupling prediction model can take full advantage of the systematic multi-time historical data and prominently possesses superior predictive performance compared with the conventionalGM(1,1)power model.It is suitable for predicting data sequences characteristics of single-peakor saturation.Thiswork makes signifigant contribution to the enrichment of grey prediction theory and the extension of its application span.