In 2004, more than 180 million personal computers (PCs) were sold worldwide. In the same year, an estimated 100 million obsolete PCs entered waste streams and were either recycled for the recovery of materials or finally disposed of. A PC may contain up to 4 g of gold and other valuable materials that can be recovered at a profit, particularly if the work is done in low-income countries. However, as is the case with almost all present-day electronic products, a PC also contains toxic substances such as lead, mercury, arsenic, cadmium, selenium, and hexavalent chromium. In many parts of the world, both formal and informal recycling industries that deal with the rapidly growing streams of Waste Electrical and Electronic Equipment (WEEE), or e-waste for short, have emerged. For the EU15 (the 15 countries constituting the EU before May 2004), the WEEE generated per capita today is between 4 and 20 kg/a (Widmer et al., in this issue). The range of uncertainty is mainly due to definitional problems, as it is typical for the entire e-waste topic. Computers are only one type of WEEE. However, given the trend towards pervasive computing, which means that more and more everyday commodities will contain microprocessors in the future, the borderlines between dclassicT electrical equipment (such as refrigerators) and electronic equipment will become blurred. One can already see today that more and more objects that used to be considered purely delectricalT are now equipped with computer chips, and thus have turned into delectronicT objects. Today, more than 98% of all programmable microprocessors are embedded in commodities that are usually not perceived as computers (e.g., household appliances and toys). Even more relevant from an environmental point of view, many commodities that until recently were considered dnon-electricT are now being equipped with microprocessors for extended functionality, or with radiofrequency identification (RFID) transponders for contactless identification (Hilty et al. 2005; Oertel et al. 2005). Both the old (device-like) and new (embedded) types of information and communi-cation technologies (ICTs) are spreading out rapidly, leaping geopolitical borders and penetrating our everyday lives across traditional categories of basic commodities. Given these trends in ICT diffusion and application, it is likely that the dissipation of valuable and toxic materials due to the distribution and disposal of electronics will continue, unless effective countermeasures are taken. The hope that the continued miniaturization of electronics, according to the so-called Moore's Law and related technological trends, will 0195-9255/$ -see front matter Deiar solve the problem in the long run is neither supported by experience nor by the expectations explicitly stated by ICT manufacturers. Experience shows that the miniaturization of devices is usually counteracted by the growing numbers of devices produced. For instance, the considerable reduction in the average physical mass of a mobile phone from over 350 g (1990) to about 80 g (2005), which corresponds to a reduction by a factor of 4.4, was accompanied by an increase in the number of subscribers, which in turn led to a rise of the total mass flow by a factor of 8.0 (data for Switzerland; Hilty et al. 2005). In every case of miniaturization in digital electronics thus far, the price per functional unit has fallen and triggered greater demand, which compensates–or even overcompensates–for the miniaturization effect in terms of mass flow. There is no evidence that this rebound effect of miniaturization will no longer apply if the visions called dpervasive computing,T dubiquitous computing,T or dambient intelligenceT become real. Quite the contrary, IBM expects that in the next 5–10 years, about 1 billion (10 9) people will be using more than a trillion (10 12) networked objects across the world. This would mean that there would be an average of 1000 dsmart objectsT per person in the richer part of the world, each containing a processor and some communication module. If we assume that the average mass of an electronic component used to make an object dsmartT is about 10 g and that such a component would be in service for about 1 year, the resulting per-capita flow of e-waste amounts to 10 kg/a. This value is on the same order of magnitude as today's e-waste in industrialized countries, as mentioned above. We can conclude that implementing the dsmart objectsT vision would not render the mass flows of electronic waste negligible; however, it will certainly change the quality and manageability of these flows. Taking other technological visions literally can even lead to dramatic results. One example is the vision of de-grainsT—very small processors that are envisioned to be used as dintelligent wall paint,T turning walls into large-scale displays and rooms into distributed computers. In a study for the Swiss Center for Technology Assessment, it was hypothetically assumed that this technology would be applied to give every inhabitant of Western Europe, North America, and Japan one dintelligent room.T Assuming further that nickel will still be used as a constituent of e-grains, it was estimated that more than 40% of the world's annual nickel production (1.2 million tonnes in the year 2000) would be required to produce the wall paint (Hilty et al 2005). This example shows that the supply of exotic raw materials could become a limiting factor for future electronics production. The temporary shortage of tantalum that occurred in 1999–2001 demonstrated this problem. Only two companies extract tantulum from the mineral coltan in the Democratic Republic of Congo and Australia. This scarcity appreciably slowed the growth of the ICT industry (e.g., in the mobile phone and games console segments) (Horvath 2002). In the project dThe Future Impact of ICT on Environmental SustainabilityT for the European Commission, a socio-economic simulation study with a time horizon running up to 2020 was carried out for three different policy scenarios (Hilty et al. 2004). Even in the scenario which assumed that environmental regulation would be put into force to internalize external costs (e.g., accounting for the externalities of extracting and processing raw materials, supplying energy, and disposing of waste), the total EU15 WEEE mass flow Editorial 432