Abstract: In order to forecast the popular state of new products in apparel market, this paper built a new kind of forecasting multiply model based on time series. Through describing the market state of any new apparel products, this paper defines three representation variables, such as acceptance ratio, trend and life cycle. Many factors such as apparel itself, market environment and customer themselves’ characters are comprehensively considered when evaluate those characters which influence market demand. All sub-functions are discussed, including day-mean index, people flow cycle features, the functions such as price influence, trend and demand on seasonal products. And, through a calculation example and computer simulation, this model is verified and the meaning of forecasting is interpreted.