The concept of a nation-wide Health Survey for Children and Adolescents has been tested over one year in a pilot study with 1,630 participating children and adolescents. A major methodological aspect of this pilot phase was to compare different variants of study sample recruitment regarding their feasibility. There was a choice between a sample originating from the population registry and a cluster sample of schools and school classes for the age group from 11 to 15 years. As a result, after weighing the advantages and disadvantages of both sampling designs, a continuous population registry sample has been recommended. The acquired knowledge and experience will be translated into reality in 2003 when the main course of the survey, which will include about 18,000 children and adolescents, will start.