A pilot survey of 200 households was undertaken to field test the survey instruments and trial the validation and calibration procedures for the Western Australian Child Health Survey (WACHS) conducted in 1992. This paper describes the background to the WACHS, the development of the instrumentation and the conduct of the pilot study. This survey aims to replicate and extend previous epidemiological surveys conducted in other countries, and to provide Australian norms for mental health morbidity in 4 to 16 year old children. The measurement of mental health was undertaken through the use of the Child Behaviour Checklist (CBCL). This screening instrument provided data on the prevalence of mental health morbidity and of specific mental disorders in 4 to 16 year olds. Its reliability and validity as a diagnostic indicator were checked through a clinical calibration technique. The pilot survey also permitted an examination of the sampling strategy adopted to ensure that the sample selected reflected "normality" in terms of expected trends and results. Modifications to the content of the questionnaires are described in light of both psychometric qualities of the data and comments from field interviewers and professionals who have examined the instrument. Finally, changes to data collection strategies are discussed.