We have recently demonstrated that genotype-specific loss of subjects may introduce bias into Mendelian Randomization studies (Rodriguez et al, American Journal of Epidemiology, in press). A lifecourse analysis of alleles and genotypes would enable us to rigorously (and comparatively with other UK cohorts) determine precise and accurate estimates of allelic and genotypic birth frequencies on which "intention to randomize" analysis could be based in Mendelian Randomization studies. We request all available primary genome-wide genotyping data (i.e. raw fluorescence data) for ALSPAC subjects to enable highly controlled (precise and accurate beyond the level of the standard automatic software) estimates to be made of genotype and allele frequencies. We also wish to analyse the raw (primary) data to determine specific technical reasons for selective loss of genotypes in individual markers. We do not intend to directly analyse genotype against phenotype data for this project; instead we will generate a data resource which will facilitate more robust genetic epidemiological analyses.