Public concern has been raised that living near a landfill site may be hazardous to health. In particular, several U.S. and U.K. studies have shown excess risk of birth anomalies in populations living near landfill sites (Dolk et al. 1998; Fielder et al. 2000; Vrijheid 2000). To investigate potential risk of adverse birth outcomes associated with landfill sites in Great Britain, investigators had access to an extensive data set of current and previously opened landfill sites provided by the environmental protection agencies in Great Britain. Data were incorporated in a GIS, resulting in a database containing 19,196 landfill sites in England, Wales, and Scotland. Detailed data on boundaries were unavailable for most sites, and therefore point locations had to be used. Site centroids were given for a majority of sites. The location of the site gateway at the time of reporting was used for the remainder. Geocoded data were supplied for landfill site locations but were of low accuracy (often rounded to 1,000 m), and area data were inadequate for most sites. Landfill site areas also changed considerably over time. Postcodes, which were used to define the location of cases and births, only approximated the place of residence. When researchers tried to intersect location of landfill(s) and residences of study subjects, they found that landfill sites are often highly clustered, so that individual postcodes may lie close to as many as 30 or more sites. Given that study subjects may be exposed to several landfill sites, distance from the nearest landfill site was not regarded as a meaningful proxy for exposure. As a compromise between the need for spatial precision and the limited accuracy of the data, a 2-km zone was constructed around each site (Figure 3), giving a resolution similar to or higher than that of previous studies (Dolk et al. 1998; Fielder et al. 2000) and at the likely limit of dispersion for landfill emissions (WHO 2001). The reference population comprised people living more than 2 km from all known landfill sites during the study period. Availability of landfills and health outcome data were restricted to the study period from 1983 to 1998.