According to estimates of June 1999, there are about 180 million Internet users world-wide: the USA and Canada alone account for 102 million, whereas in Europe 43 million people are online (Nua Internet Surveys, http://www.nua.ie/). These numbers prove that the Internet cannot be considered a fringe phenomenon. Even if at present, the majority of the population does not have access to the Internet, we must ask ourselves whether it is already viable to use the opportunities of the net for acquiring sociological data – i.e. by conducting surveys via the Internet. The advantages of such a venture are obvious. In comparison to postal surveys, the financial expenditure of surveys on the Internet is smaller, and surveys can be conducted within short periods with an extremely high number of cases. However, these advantages are confronted by serious methodical problems. Above all, difficulties in determining population and random samples need to be mentioned. Primarily, these problems concern World Wide Web surveys, i.e. surveys in which anyone with Internet access can participate, and where a sample selection cannot be compiled according to a well-defined design. However, today there are already numerous applications in which the advantages of the net can be used in accordance with the usual methodological standards. These include, for example, employee surveys in globally dispersed and networked corporations, surveys on particular populations (cf. amongst others Schaefer & Dillman, 1998) or experts, as well as surveys of registered Internet users in defined areas – in other words, surveys in areas where the respective populations are clearly defined, and either total censuses or random samples are possible. Such methodically harmless surveys are not the subject of this contribution. In the following section the difficulties involved in using data acquired in this manner to make general statements on the Internet user population will be pointed out in a short outline based on what could be called the most famous World Wide Web survey, the GVU User Surveys.