When making projections in Environmental Outlooks, in estimating the future environmental costs RIVM considers these costs as being constant through time. However, in reality these costs are not constant, but they decrease as a result of scale- and learning effects. This report presents a method to determine the decrease of costs during a period of time. In doing this, empirical data about environmental technologies is collected. Five clusters are composed, and per cluster a formula is derived that describes the cost decrease through time. These formulas are fitted with the help of a statistical package. Using RIVM's database of environmental technologies, an exercise is done calculating the environmental costs by attaching each of the 800 technologies to one of the five clusters. The conclusion is that by considering costs to be constant the total environmental costs are overestimated. The exercise shows that this overestimation is about 10%. Because of the limited set of empirical data, the clustering of the technologies and the arbitrary attachment to the clusters the results of this study should be considered as indicative.