Today's stocks are tomorrow's emissions and waste flows. As a result of the time lag introduced by the buffering function of the stock of materials and products in society environmental problem flows which seem to be under control can easily rebound. In this paper an example is given of how signal processing can be used in dynamic Substance Flow Analysis for estimating the future generation of waste and emissions from present societal stocks. An approach is outlined to estimate the outflow of waste products from stocks on the basis of assumptions on the shape of the distribution describing the inflow of new products, the average life span of the products, and the life-span distribution. To exemplify the approach we used a theoretical case of PVC in Sweden. It was found that the delaying mechanisms of the stocks can make the outcome counterintuitive. Furthermore, the chosen shape of the input-distribution function has the most influence on the predicted outflows, especially in the case of possible fashion-type (exponentially increasing) markets. The choice of the shape of the inflow distribution could, therefore, be based on qualitative knowledge of the market of the different products. The life-span distribution appears to have a more subtle influence on the height of the peaks and the time that they occur. So far only a normal distribution has been considered; more research is recommended into other types of distribution.